See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/254401648

HIV/AIDS in Kenya: A Review of Research and Policy Issues

Article

CITATIONS READS

8 2,893

3 authors, including:

Diana Kimani Germano Mwiga Mwabu

University of Nairobi University of Nairobi
9 PUBLICATIONS   58 CITATIONS    73 PUBLICATIONS   1,731 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Equity in health and healthcare in Malawi: Analysis of trends View project

Entrepreneurship as a mechanism to address youth unemployment and poverty View project

All content following this page was uploaded by Diana Kimani on 08 May 2014.

The user has requested enhancement of the downloaded file.

HIV/AIDS in Kenya: A Review of
Research and Policy Issues

Robert K. Nyaga
Diana Njeri Kimani
Germano Mwabu
Mwangi S. Kimenyi

Social Sector Division

Kenya Institute for Public Policy
Research and Analysis

KIPPRA Discussion Paper No. 38

June 2004
HIV/AIDS in Kenya: A review of research and policy issues

KIPPRA IN BRIEF
The Kenya Institute for Public Policy Research and Analysis (KIPPRA)
is an autonomous institute whose primary mission is to conduct public
policy research, leading to policy advice. KIPPRA’s mission is to produce
consistently high-quality analysis of key issues of public policy and to
contribute to the achievement of national long-term development
objectives by positively influencing the decision-making process. These
goals are met through effective dissemination of recommendations
resulting from analysis and by training policy analysts in the public
sector. KIPPRA therefore produces a body of well-researched and
documented information on public policy, and in the process assists in
formulating long-term strategic perspectives. KIPPRA serves as a
centralized source from which the government and the private sector
may obtain information and advice on public policy issues.

Published 2004
© Kenya Institute for Public Policy Research and Analysis
Bishops Garden Towers, Bishops Road
PO Box 56445, Nairobi, Kenya
tel: +254 20 2719933/4; fax: +254 20 2719951
email: admin@kippra.or.ke
website: http://www.kippra.org
ISBN 9966 949 63 1

The Discussion Paper Series disseminates results and reflections from

ongoing research activities of the institute’s programmes. The papers
are internally refereed and are disseminated to inform and invoke debate
on policy issues. Opinions expressed in the papers are entirely those of
the authors and do not necessarily reflect the views of the Institute.

KIPPRA acknowledges generous support from the European Union

(EU), the African Capacity Building Foundation (ACBF), the United
States Agency for International Development (USAID), the Department
for International Development of the United Kingdom (DfID) and the
Government of Kenya (GoK).

2
Abstract
This paper examines the magnitude, distribution and causes of HIV/
AIDS in Kenya, including responses to fight the disease. An account is
also provided of theoretical and empirical economics research
approaches used in analysing the impact of HIV/AIDS. On average,
statistics show that HIV/AIDS prevalence rates in urban areas are higher
than in rural areas, and that HIV/AIDS is imposing heavy costs on the
economy, primarily through increased medical care expenditure and
labour losses. There are no strong indications that the epidemic is
declining, casting doubts on the success of past anti-AIDS efforts. The
true impact of awareness-creation campaigns in reducing new infections
is still uncertain despite that most funds for intervention are channelled
to these efforts. Medical treatment of AIDS patients, which can prolong
life and prevent loss of productive labour, has received little attention
or funding.

This study suggests a partial policy shift from awareness campaigns to

prevention, early diagnosis, treatment, and economic and social support
for those infected and affected. To forestall momentous economic losses,
we recommend an incentive mechanism that promotes three important
inputs to the national AIDS management programme:

• Data on early-diagnosis and follow-up programmes to allow for

animation of case-specific interventions, with voluntary
counselling and testing (VCT) forming the backbone for medical,
economic and social support regimes;

• Collection of incidence and general prevalence data to serve as

impact indicators and for impact analysis and long-term planning;

• Favourable and sustainable linkage of all awareness campaigns,

VCT drives and follow-up strategies, treatment regimes and social
and economic support programmes to the Poverty Reduction
Strategy Papers (PRSP) and the budgeting process, i.e. Medium
Term Expenditure Framework (MTEF), with greater commitment
of government allocations to these programmes rather than
dependency on donor funding.

3
HIV/AIDS in Kenya: A review of research and policy issues

This Discussion Paper is produced under the Umbrella Project for Improving
the Enabling Environment for Businesses in Kenya. The aim of the Project is to
improve the policy, legal, and regulatory environment for businesses. The
Project has three components. The Simplifying the Regulatory
Environment for Business (SREB) component involves research on
constraints to operation of business by the private sector in Kenya. The
Private Sector Advocacy component assists the private sector in advocating
for reforms that create a favourable environment for business and
investment. The Capacity Building component aims to build capacity in
line ministries and regulatory agencies to respond to reform proposals
made by the private sector and other stakeholders. KIPPRA implements
the first and third components while the Kenya Private Sector Alliance
implements the advocacy component. The Project is funded by the British
Department for International Development (DfID).

4
Acronyms
ACP National AIDS Control Programme, Uganda
AIC AIDS Information Centre, Uganda
AIDSCAP AIDS Control and Prevention Project
APS AIDS Programme Secretariat
ARV Antiretrovirals
CBS Central Bureau of Statistics
CGE Computable General Equilibrium
COI Cost of Illness
FHI Family Health International
IEA Institute of Economic Affairs
MTEF Medium Term Expenditure Framework
NACC National Aids Control Council, Kenya
NAPCC National AIDS Prevention and Control Committee,
Thailand
NASCOP National AIDS/STD Control Programme
PRSP Poverty Reduction Strategy Paper
SID Society for International Development
STD Sexually Transmitted Disease
STI Sexually Transmitted Infection
TASO The AIDS Support Organization
UAC Uganda AIDS Commission
VCT Voluntary Counselling and Testing
WHO World Health Organization
WTP Willingness to Pay

5
HIV/AIDS in Kenya: A review of research and policy issues

6
Table of Contents

1. Introduction ......................................................................... 1

2. HIV/AIDS in Kenya ........................................................... 4

2.1 Determinants of HIV/AIDS prevalence ................... 6

2.2 Impact of HIV/AIDS in Kenya ................................ 15

2.3 Government response to the AIDS epidemic ......... 20

2.4 Experience of other countries with HIV/AIDS ..... 24

3. Theoretical Issues and Empirical Evidence ................... 29

3.1 Micro and meso-level effects .................................... 29

3.2 Macro-level effects ..................................................... 36

4 AIDS Research Methodologies in Economics ............... 41

5. A Case for Further Research ............................................ 48

6. Priorities and Policy Issues .............................................. 52

References ........................................................................... 56

Appendix ............................................................................ 60

7
HIV/AIDS in Kenya: A review of research and policy issues

8
1. Introduction

Since its emergence in the early 1980s, HIV/AIDS has spread at an

alarming rate worldwide with the number of new infections rising each
year. Africa continues to bear the brunt of the scourge. Globally 42
million people are infected with HIV/AIDS, 29.4 million of them in
sub-Saharan Africa (Appendix table A1). There were 3.5 million new
HIV/AIDS infections in sub-Saharan Africa in 2002 out of the 5 million
worldwide (Appendix table A2). Furthermore, majority of the more than
21.8 million people who have died from HIV/AIDS since the start of
the epidemic from Africa. In addition, AIDS deaths have left over 8
million orphans in sub-Saharan Africa. These high rates of HIV/AIDS
incidence make the disease the leading cause of death in Africa (WHO/
UNAIDS, 2002).

HIV infection rates in Kenya were low in the 1980s. However, by the
end of 1998, almost 14% of Kenya’s adult population was living with
the HIV virus, or about 2.1 million people (WHO/UNAIDS, 2002).
According to the Kenya Economic Survey for 2003 (GoK, 2003). HIV
prevalence rates have however declined from about 13% in 2000 to 10.2%
in 2002. This is a tremendous decline, which is attributed to vigorous
advocacy campaigns against the disease by the government and other
organizations.1

Kenya ranks ninth in Africa in AIDS prevalence, with Botswana leading

with an adult infection rate of 35.8%, followed by Swaziland and
Zimbabwe with rates of 25.25% and 25.06%, respectively (Loenwenson
and Whiteside, 2001).

1
The adult prevalence data, unlike incidence studies, may potentially be
misleading. For instance, falling prevalence rates do not indicate that new
infections are declining, which may be of significance in policy.

9
HIV/AIDS in Kenya: A review of research and policy issues

Whether HIV/AIDS incidence rises or falls in the future does not

fundamentally change the understanding that the disease poses grave
health and developmental problems in Kenya. Although the HIV/AIDS
impact is difficult to measure, it has been shown that growth in
economies badly hit by the epidemic may be 25% lower than it would
have been in the absence of AIDS. Studies show that the macroeconomic
impact of AIDS in terms of declining GDP growth rate is about 0.3 to
1.5% in Africa (Bonnel, 2000; Arndt and Lewis, 2000). The disease affects
individuals and households (the human capital stock) resulting in
serious negative effects on almost all sectors of the economy, including
household and individual welfare.

Despite the many efforts to fight AIDS and the rising awareness of the
disease, the epidemic continues to claim lives while imposing heavy
costs on the Kenyan economy. Awareness creation campaigns have not
developed the tempo necessary to reverse the spread and cushion the
economy against the impact of the epidemic. Further, reliable economic
research on the actual cost of the disease is lacking, which means that
policy on the disease is based on speculative and non-objective data.
Much of the research in Kenya is either sector-specific case studies or
surveys of specific risk groups, including medical and clinical studies.
Some of these may be regarded as doomsday-scenario studies designed
to raise awareness or justify funding for the fight against AIDS.2 This
could have been necessitated by the inherent nature of the AIDS scourge
and data problems, which have hindered more scientific and objective
research on the economic aspects of the disease. Reliable incidence3 data
have been difficult to collect owing to the stigma attached to HIV/AIDS

2
Lack of objectivity in research on AIDS may propagate stigmatization against
those infected, which goes against the desired public objectives.
3
The Centre for Disease Control (CDC), Central Bureau of Statistics (CBS) and
National AIDS/STD Control Programme (NASCOP) are conducting a survey
based on actual HIV tests and targeting 8000 households.

10
 Introduction

infection. Basically, reliability of the sentinel data used to estimate adult

HIV prevalence is questionable on many grounds, even with improved
demographic and data-collection skills.

Descriptive studies have been conducted on the effects of AIDS on

various sectors, households and individuals, but these effects are rarely
quantified. The impact of the pandemic on the whole economy and on
macro indicators such as economic growth, interest rates, levels of
investment and human capital stock has not been assessed in Kenya.
However, it is possible to quantify the AIDS impact on these and other
indicators using various economy-wide simulation models alongside
econometric models. This paper attempts to develop both theoretical
and empirical frameworks for such an effort. The paper also briefly
reviews the AIDS situation in Kenya and the economic research
methodologies used in AIDS research, and identifies areas for research.
The paper also discusses various methods and strategies used to fight
AIDS. These include the new thinking about AIDS interventions and
associated factors concerning funding for interventions, VCT services,
and access to antiretrovirals (ARVs) and treatment for opportunistic
diseases and other assistance to those infected and affected, all which
will help stem possible collapse of human capital and productivity.

The following section presents an overview of HIV/AIDS in Kenya.

Section 3 discusses theoretical and empirical accounts of AIDS research,
while Section 4 reviews various research methodologies. Sections 5 and
6 identify areas for further research and provide a policy relevant
discussion.

11
HIV/AIDS in Kenya: A review of research and policy issues

2. HIV/AIDS in Kenya

Since the emergence of HIV/AIDS in Kenya in 1984, the epidemic has

spread unabated. Fairly reliable estimation techniques put the number
of adults and children living with HIV/AIDS at about 2.5 million by
year 2001 according to WHO/UNAIDS (2002). Adults aged 15-49
account for 2.3 million of this number, and children account for 0.22
million (table 1). The table also shows that over half of all those infected
are women.

Table 1: Estimates of HIV/AIDS situation by 2001

Group Number
infected

Adults and children living with HIV/AIDS 2,520,000

Adults living with HIV/AIDS (15-49) 2,300,000
Women living with HIV/AIDS (15-49) 1,400,000
Children (0-15) 220,000
Deaths in 2001 190,000
Number of orphans below 15 years old = 890,000

Source: WHO (2002); UNAIDS (2002)

HIV prevalence among adults in Kenya was expected to increase from

5% in 1990 to 14% at the end of 1998, pointing to possible negative effects
on the countries’ human capital base. However, HIV prevalence rates
declined nationally from 13.4% in year 2000 to 10.2% in year 2002 (table
2). The rates are higher in urban centres than in the rural areas. Urban
infection rates stood at 17.5% in year 2000 compared with 13.5% for
rural areas. The rates have declined in both regions. The slight decrease
in prevalence rates in year 2000 is mainly due to high mortality rates
(GoK 2001a) rather than low infection. However, further decreases to

12
 HIV/AIDS in Kenya

10.2% in 2002 could be attributed to intervention programmes against

HIV/AIDS.

Table 2: Prevalence rates for 1996-2003

Prevalence 1996 1997 1998 1999 2000 2001 2002 2003

Urban 16.3 16.9 18.1 17.8 17.5 17.0 16.5 10.0

Rural 11.0 11.9 13.0 13.0 13.0 13.0 12.5 5.6
National 11.9 12.8 13.9 13.5 13.4 13.0 10.2 6.7
Source: Government of Kenya (2001a, 2003)

For the year 2002, Nyanza Province had the highest prevalence rate of
23%, followed by Eastern and Rift Valley provinces each with 19%.
Nairobi had a prevalence rate of 9%, though it has the highest rate among
urban areas. North Eastern Province has the lowest rural and urban
prevalence rates (Appendix table A3).

Sentinel data for sexually transmitted illnesses (STIs) for 1990-2001 show
that prevalence rates have been rising among all categories of marital
status. Prevalence is highest among the widowed (implying that their
spouses most likely died from AIDS-related ailments) and among people
who are separated, followed by people in polygamous unions. It is
lowest among single people and those in monogamous unions
(Appendix table A4). These variations in rates between monogamous
and polygamous unions may be explained by the fact that polygamous
unions expose more people to infection from one person, whereas in
monogamous unions the spread is limited to two people. The high rates
among separated people are likely to be associated with the likelihood
that they may engage in risky sexual behaviour following disruption
of their normal life and as a way of consoling themselves.

Gender is an important factor in AIDS incidence. Young women aged

15-19 are five times more likely than men in this group to be infected
with HIV, and women aged 20-24 three times more likely than men in

13
HIV/AIDS in Kenya: A review of research and policy issues

the same age group. Various reasons, though devoid of reliable research
evidence, have been put forth as to why young women are more at risk
than men. These include gender-specific biological factors, the tendency
for younger women to have sexual relationships with older men, and
the poor economic status of women. Statistics (UNAIDS/WHO, 2000)
show that commercial sex workers are the hardest hit group by AIDS,
and prevalence among this group has been rising since the first cases of
HIV were reported in Kenya in 1984. In Nairobi, HIV prevalence rates
among sex workers had reached 62% in 1985 and increased to 86% by
1992 (UNAIDS/WHO, 2000)

The large number of AIDS orphans provides a grim picture of the AIDS
epidemic. Since the epidemic was first reported, the number of children
who have lost either their mother or both parents to AIDS when they
were less than 15 years old had reached about 890,000 in 2001 (table 1).
This poses great challenges for the government, since these children
need care and schooling to grow into productive citizens.

2.1 Determinants of HIV/AIDS prevalence

There is no consensus on either the major causes of HIV in Africa or

why Africa accounts for the overwhelming share of HIV/AIDS cases
in the world. The main modes of transmission of HIV are sexual
intercourse, re-use of contaminated syringes by drug users, mother to
child infection through birth or nursing, re-use of needles in medical
settings, and transfusion of contaminated blood or blood products. HIV
cannot be transmitted by a sneeze, a handshake or other casual contact.4
However, it is believed that heterosexual sex, compounded by the unsafe
sex practices dominant in Africa, is the leading transmission mechanism

4
HIV can be isolated from the saliva of an infected person. Although there are
a handful of cases of transmission through oral sex, there are no confirmed
cases of transmission via saliva alone.

14
 HIV/AIDS in Kenya

of HIV. This may explain why most anti-AIDS campaigns advocate

sexual behaviour change and use of safe-sex options such as condoms.
However, safe sex options are not widespread in Africa, where condom
use is still ‘taboo’ and where the poor might not afford to spend Ksh 10
on a condom pack.5

A survey carried out by the Government of Kenya in collaboration with

UNICEF covering the whole country in year 2000 (i.e., Multiple
Indicators Cluster Survey (MICS) Report of 2000) shows that knowledge
about the key means of prevention is low (Appendix table A5). Only
52% of women and 66% of men said they knew that a condom may be
used to prevent transmission of HIV. More urban women than rural
women are aware of the use of condoms in AIDS prevention. But the
opposite is true for men. At the same time, both men and women claimed
knowledge that faithfulness to one’s partner or abstaining from sex was
an option to prevent AIDS transmission. But knowledge of prevention
mechanisms did not differ much with age, although it rose with
education. For instance, only 39% of women without education were
aware of the use of condoms for HIV prevention compared with 62% of
those with secondary school education or higher. Similar trends were
observed for other prevention methods. Some 83.3% of women and 93%
of men were aware of at least one way of preventing HIV infection
(GoK/UNICEF, 2003). The process through which knowledge translates
into adoption of safe behaviour is complex and cannot be deduced from
these statistics.

Factors that may drive the spread indirectly include lack of treatment
for people with HIV/AIDS or lack of access to HIV/AIDS testing
facilities. It is argued that individuals who test positive have a high
likelihood of avoiding unprotected sex, and those who test negative

5
Alexander Cockburn. ‘Why Africans get AIDS? Creators Syndicate-
www.workingforchange.com

15
HIV/AIDS in Kenya: A review of research and policy issues

have an incentive to stay that way. 6 GoK/UNICEF (2003) gives

percentages of men and women who knew of a place to take a HIV test
and those who have ever taken a test. Women are more likely than men
to test for HIV even though men are more knowledgeable about facilities
for testing. About 64% of women know of a testing place but only 10.4%
have been tested; for men, this is 76% and 8.2%, respectively (Appendix
table A6). Further, fewer rural women and men have been tested. For
both women and men, the higher the level of education the higher the
numbers that have been tested.

Several factors are thought to influence the spread of AIDS in Kenya,

though none can be said to be dominant.

Cultural and gender factors

In Kenya, marriage increases exposure to HIV/AIDS, especially if one

partner is not faithful. Women are disproportionately more vulnerable
to infection. Primarily for them, a marriage may serve as a sure infection
trap if the man is not faithful. This is made worse by cultural and socio-
economic imbalances that favour men and leave women powerless in
most decisions. Women in this context have little control over their own
sexual behaviour let alone on that of their partner and cannot demand
that their partner use preventive means to reduce chances of infection.
Instances of physical violence are common in Kenya, including sexual
violence.

An unfaithful partner may endanger the other, but the problems faced
by women in general include exposure to infection as they give care to
infected people without sufficient protection or information.

6
These are behavioural predictions that do not work independently of a maze
of other factors; therefore, the direction of behaviour, especially in the long
term, may be hard to predict a priori.

16
 HIV/AIDS in Kenya

The desire by Kenyan women for many children accentuates the HIV/
AIDS problem in the country, since it discourages use of preventive
options such as condoms. Other risky cultural factors include teenage
marriage and forced marriage of young girls to older men. Gender
inequities in distribution of wealth and power relations give rise to
asymmetric age matching, fostering sexual relations between young
women and older men who are likely to be infected (IEA/SID, 2001).
Widow inheritance is a risky cultural practice accounting for high HIV/
AIDS infection rates. This rite is reinforced by taboos, superstition and
fear of bad omen (IEA/SID, 2001).

Studies (NASCOP 1998, 1999), though inconclusive, show some effect

of male circumcision in reducing the chance of infection. This is a
controversial issue that cannot fully be settled through studies. In studies
at a special treatment clinic in Nairobi on male circumcision and the
risk of STIs and HIV, sero-conversion rates were 2.5% for circumcised
and 29% for uncircumcised men. For men with genital ulcer disease,
the rates were 13.4% for circumcised and 52.6% for uncircumcised
(NASCOP, 1999). Female partners of uncircumcised HIV-positive men
also run higher risk of infection. However, being circumcised is not a
shield from HIV infection, since in some communities circumcision
rituals subtly encourage new initiates to seek sexual encounters as proof
of their manhood. In other communities, women are encouraged to
secretly consider bearing at least one child from an extramarital affair,
to spread the genes of their children7.

The collapse of positive traditional cultural value systems takes some

of the blame for the spread of HIV/AIDS. For instance, replacement of
traditional marriage practices with modern value systems has given
rise to prevalence of cohabitation in both rural and urban Kenya. This

7
These complex cultural norms and values are not always in the public domain
and therefore are hard to stamp out.

17
HIV/AIDS in Kenya: A review of research and policy issues

also includes degradation of traditional support systems, especially in

urban areas, and lack of clear cultural identity among the youth, who
are exposed to both traditional and western influences.8 Inadequate
communication and awareness campaigns about HIV/AIDS and its
prevention have also been responsible for the spread of HIV/AIDS9.
For instance, parents shy away from discussing sex issues with their
children. Also, members of parliament, who play a crucial role in policy
formulation and advocacy and who can be used to sensitise their
constituents on the issues of AIDS are still shying away from discussing
issues of HIV/AIDS. A media opinion poll, in which 3,000 people
throughout the country were interviewed shows that religious bodies
have played a insignificant role in creating awareness about AIDS.
Churches and mosques were only responsible for creating awareness
among 4.7 per cent of respondents. An analysis by age shows that 15.5
per cent of teenagers were not aware of AIDS (Daily Nation, 2001). Other
factors that influence HIV transmission are age at first sexual intercourse,
number of partners, drug use, among others (WHO/UNAIDS, 2002).

Biological factors

Some biological characteristics particularly for women increase the risk

of HIV infection. Transmission of the virus from men to women is at
least four times more likely than from women to men. Women are more
vulnerable to infection because the features permit greater mucosal
(surface) exposure to seminal fluids. This is coupled with prevalence of
non-consensual sex, unprotected sex, and high-risk sex partners. Sex

8
Thanks to the media and particularly to the new information and
communication technologies, the youth have to cope with vast and diverse
information on sexual behaviour.
9
We see later how unstable information-based interventions such as awareness
campaigns are in predicting effect on infections.

18
 HIV/AIDS in Kenya

among younger women and forced sex may cause micro lesions that
raise efficiency of infection. Biological factors also influence the spread
of the epidemic by increasing or decreasing susceptibility to the virus
and hastening the progression of infection to disease and likely death.
Such factors include the presence of sexually transmitted diseases, viral
loads and type of HIV. Such factors should inform response and
prevention strategies, which may include seeking use of microbicides
in the case of women and antiretroviral therapy, among others.

HIV may be transmitted from mother to child during pregnancy, birth

or after birth. This transmission pathway is common but can be
minimized using AZT or a combination of antiretroviral drugs. The
risk of mother-to-child transmission is increased if the mother is at an
advanced stage of HIV, uses drugs or suffers severe inflammation of
foetal membranes during birth, or if the period between membrane
rapture and delivery is prolonged.

Economic factors

Economic factors play a pivotal role in the spread and control of HIV/
AIDS. To start with, globalization and migration have facilitated free
movement of people, goods and services across countries, which has
led to separation of families for long periods, predisposing people to
risky behaviour. In the developing world, increased mobility in search
of jobs, rural-urban migration and industrialization have been identified
as major factors in the spread of HIV. Mobile workers who are the most
vulnerable to HIV/AIDS include those in the transport, fishing and
tourism industries and migrant labourers in mines, oil fields, roads and
dam projects. Major development projects requiring large migrant
labour have increased the rate of the spread of HIV/AIDS.

The economic aspect of the spread of AIDS is not restricted to

globalization and migration, but includes indebtedness and high costs

19
HIV/AIDS in Kenya: A review of research and policy issues

of HIV/AIDS-related drugs and equipment. For a long time, AIDS drugs

such as AZT and ARVs, which delay progression of infection to fully
blown AIDS have been out of reach of most patients in sub-Saharan
Africa. The drugs basically catered for the rich who could afford. A
GlaxoSmithKline’s (GSK) HIV drug treatment cost USD 12,000 a year,
compared with per capita health spending of USD 23 in low income
countries. Intense lobbying across the world has helped lower the cost
of anti-AIDS regimen considerably. GSK’s HIV drug fell to USD 727
per year in 2003 (Bosley, 2003). However, this was still more than double
the cost of equivalent generic drugs such Triomune, a three-drug
cocktail, manufactured by Cipla of India, which costs $304 per year
(Boseley, 2003). Though willing to lower their prices, the hands of the
main pharmaceutical companies are tied by the profit motive. Besides,
these firms cannot make cheap generics like the generic companies due
to patent restrictions. The plausible route therefore has been for them
to argue that the UN buy ARVs in bulk and distribute them to poor and
needy countries.

Drugs that prevent mother-to-child transmission of HIV are still in acute

shortage in most developing countries. Supporting infrastructure and
equipment such as HIV testing kits, which may have no cheaper
alternatives, are also not easily available or are inappropriate for Kenya.
Shortage of items like blood transfusion equipment, syringes and
equipment for medical procedures may pose transmission risks. In
countries like Kenya therefore demand for drug therapy within the
average person’s purchasing power outstrips supply.

Some of the poor, AIDS-ravaged nations are also the most indebted.
The Heavily Indebted Poor Countries (HIPC) initiative created by G8
countries provides to waive the debt of some 33 poorest countries, with
the reductions in debt commitment used in poverty and HIV/AIDS
reduction. However, this misses the point because countries that qualify

20
 HIV/AIDS in Kenya

for the waiver are not necessarily the most affected by the scourge. For
example, although the scourge has heavily affected countries such as
Kenya, Botswana and South Africa, these countries are not among the
33 poorest countries. They would therefore not qualify for the waiver.

Other problems that inhibit control and prevention of AIDS are poor
access to medical care, inefficiency in delivery of medical services, and
inadequate training of medical staff to handle AIDS patients.

Provision of anti-AIDS drugs is a form of public service with high

positive externalities, which cannot be left to the markets and cannot
be delayed without incurring heavy negative externalities.

Stigmatization

Stigmatization of people with HIV/AIDS leads to their discrimination

and rejection at home, at the workplace and in the public. Workers
infected with HIV/AIDS may be dismissed from employment or denied
common benefits. In the family and community environment,
discrimination takes the form of social ostracism and exclusion from
the usual family or community networks. In any of the cases, gender
biases become accentuated, as infected women face more rejection than
men. The GoK/UNICEF (2003) report documents some aspects of
discrimination perceptions (Appendix table A7). For example, some 39%
of women and 51% of men believed that a “teacher with HIV should
not teach” (GoK/UNICEF, 2003). Also, 33% and 42% of women and
men, respectively, said they would not buy food from a person with
HIV/AIDS. This shows that discrimination is marked among both
genders, but is higher among men than women. This also points to
clear misconceptions about how HIV is transmitted.

Stigmatization has had a big role in discouraging individuals from going

public about their HIV status, which has been associated with avoidance

21
HIV/AIDS in Kenya: A review of research and policy issues

of risky sexual behaviour. Also, going public about HIV status makes
one avoid risky behaviour because one is aware that people know that
you are HIV+. In effect, stigmatization may cause dejected HIV positive
people to strike back with vengeance, deliberately spreading the virus.
Therefore, the costs of stigmatization go beyond psychological and
emotional anguish of the infected person; they include the cost of new
infections that would have otherwise been prevented if the community
were more supportive and caring. It is important to appreciate the
difficulty inherent in fighting stigmatization owing to its complex nature
and that it is borne by AIDS patients. Many positive messages ostensibly
designed to reduce stigmatization or raise awareness about HIV/AIDS
and protection for those infected by the disease may subtly chide or
torment those infected with HIV/AIDS.

Political and religious factors

There is a marked political dimension to the HIV/AIDS scourge. This

mainly relates to disagreement over origins and causes of AIDS and
how to tackle it. Sexual intercourse, an epitome of human privacy, is
the main avenue through which AIDS is transmitted. Discussion about
the role of sexual intercourse in HIV/AIDS transmission causes deep
passions in both political and religious arenas. Failure by political and
religious leaders to acknowledge the existence of the disease or to allow
use of available preventive methods to curb the spread of the scourge
attests to this, and has had bad effects on efforts to stop the spread of
AIDS. For example, the Kenya government took long to concede to the
fact that the HIV/AIDS pandemic was actually an economic, social and
security problem. Religious leaders blame it on immorality and consider
it a divine punishment for the promiscuous. Such views from opinion
leaders influence the choice of prevention and intervention mechanisms
for the AIDS problem, notwithstanding politics associated with AIDS
drugs and funding for anti-AIDS efforts.

22
 HIV/AIDS in Kenya

2.2 Impact of HIV/AIDS in Kenya

Available research (Nalo and Aoko, 1993; Forsythe et al, 1993; Forsythe
and Roberts, 1995) though not exhaustive casts some light on the
debilitating effects of AIDS on the Kenyan economy. In this section, we
review evidence and studies on the effect of HIV/AIDS on various
sectors of the economy. However, we note that much of this evidence is
based on sectoral surveys, purposive studies and qualitative or non-
quantitative analytical studies.

Health sector

HIV/AIDS affects the health system by undermining its capacity to

perform, and by eroding the quality of care and the state of health
facilities. AIDS also significantly affects health-related expenditures of
households, private organizations and public health providers
(Loenwenson and Whiteside 2001). Healthcare services for AIDS
patients such as hospitalization and antiretroviral or AZT drugs in
Kenya are inadequate given that the annual per capita public health
expenditure in Kenya is only Ksh 216. This is a far cry from between
Ksh 23,000 and Ksh 43,000 per month considered sufficient for effective
management of an AIDS patient (IEA/SID, 2001).

Another problem affecting the health sector, especially the public

healthcare system, is distortion of the health referral system as demand
for quality care and desire for anonymity force AIDS patients to bypass
primary care facilities and choose more expensive tertiary institutions.
Also, AIDS has led to overcrowding in public health facilities with
patients with AIDS-related illness occupying most beds. In 1992, 15%
of hospital beds were occupied by AIDS patients; this rose to about
50% by 2000 (IEA/SID, 2001). The growing number of AIDS cases will
aggravate this problem.

23
HIV/AIDS in Kenya: A review of research and policy issues

Apart from the direct health impacts, the epidemic also affects the health
sector ’s human capital. It reduces the productivity of the health
workforce through infection, illness and death of health workers or
members of their families. Studies in Thika and Gatundu districts10
showed that doctors were highly vulnerable to HIV infection because
safety standards are poor and there is poor access to post-exposure
prophylaxis using ARVs. The reduction in the health sector workforce
due to absenteeism and death compounds the already severe human
resources problems in the sector.

Education sector

Though education is crucial to national development, income growth,

labour productivity and creation of employment, it is severely distressed
by the AIDS scourge. First, AIDS affects education demand by lowering
school enrolment as AIDS orphans increase and as more children get
infected with AIDS. The rise in child labour and child-headed
households following the death or incapacitation of the breadwinner
prevents such children from attending school (GoK, 2001a). Second, it
affects education supply through loss of teaching staff and increasing
teacher absenteeism. Third, it affects the quality of teaching and disrupts
the organizational stability of schools with the loss of key education
officials and headteachers.

Agriculture

Agriculture employs a large segment (80%) of the labour force and

accounts for 70% of Kenya’s export earnings; it is a major portion of the
GDP. As the epidemic progresses, the agricultural sector is forced to
adjust to the decrease in adult labour on farms (Rugalema et al, 1998).
This means productivity will decline and knowledge about indigenous

10
Otieno J. and M. Mwaniki (Daily Nation, 23 September 2003). “Study
highlights risks facing doctors in district hospitals.”

24
 HIV/AIDS in Kenya

farming methods will be lost. The consequences on food security are

obvious.

Agricultural companies face steadily rising costs and losses in profits

due to loss of workers, absenteeism and increased medical and funeral
costs. Labour-intensive agricultural enterprises such as sugar estates,
which depend heavily on outgrowers, lose reliable workers and farmers
to the epidemic. Consequently, they have to face a decrease in the
quantity and quality of inputs, leading to a poor recovery ratio in sugar
production (Rugalema et al, 1998).

Industry, business and markets

As the HIV/AIDS epidemic spreads in Kenya, it seems to pose the

greatest threat to the health and well-being of employees. In some
companies such as the agro-industries of the ‘sugar belt’ in Kisumu
District, roughly one in three employees is HIV positive (NASCOP,
1998). Direct and indirect costs imposed on industries include lost work
hours from worker absenteeism and reduced work effort due to illness.
Evidence shows that between 1995 and 1997, one agro-industry lost a
total of 8007 labour days due to illness of its employees (IEA/SID, 2001).

Other costs are associated with loss of experience or special skills when
highly trained and experienced employees die or are incapacitated. For
instance, it takes several years to train a fresh graduate to become a
sugar engineer. Loss of such skilled and highly trained persons results
in setbacks in sugar processing and production, and in other highly
labour-intensive agro-industries.

Loss of workers necessitates recruiting of new employees. This has cost

implications relating to advertising, selection and training. In many
cases, each death forces the remaining workers to share the tasks
previously handled by the dead colleague. In sugar factories, for
example, the average working time increased from 1 workday in the

25
HIV/AIDS in Kenya: A review of research and policy issues

late 1980s to 1.6 workdays in 1998. This may lead to work stress and
extra costs in terms of overtime payments (Rugalema et al, 1998).

Both morbidity and mortality compromise efficient planning,

management and use of human resources in an industry by making it
difficult for managers and supervisors to budget time and other
resources for a workforce. All these result in reduced labour productivity
and affect output and profits in labour-intensive industries. Frequent
illness of experienced staff and the consequent long working hours of
non-morbid employees partly explain the consistently declining cane-
to-sugar recovery ratio in the sugar estates (Rugalema et al, 1998).

Apart from productivity losses, firms face medical and funeral costs
upfront, with serious effects on profits. Data obtained from company
clinics in the sugar belt of western Kenya show that medical expenses
have increased 10-fold in the last 10 years due to increasing number of
AIDS cases (IEA/SID, 2001). Employers spend an average Ksh 45,000
in funeral costs for each deceased employee. In addition, about 6 other
employees get leave and transport to attend the funeral (IEA/SID, 2001).

A study of some Kenyan firms (IEA/SID, 2001) showed that AIDS could
increase labour costs by 15% by 2005; on average, the annual costs due
to AIDS could reach Ksh 4.3 million per business. The effect of AIDS on
small firms and informal firms has not been studied, but it is not
impossible to predict. Sole proprietorship firms are likely to simply
collapse with infection and eventual progression to death of the owner.

Individuals and households

For married women, inheritance patterns, economic subordination and

the absence of restraint on the number of sexual partners a man may
have all weaken marriage as a protective institution against HIV
transmission (NACC, 2002). While these dimensions of gender inequity
are recognized as factors in sustaining the epidemic, many responses

26
 HIV/AIDS in Kenya

to AIDS do not adequately address unequal power relations that increase

women’s vulnerability, nor have there been studies to explore the special
impact on men.

The problem of AIDS orphans clearly stands out as a major consequence

of the increased AIDS cases in households. The number of AIDS orphans
grew from an estimated 300,000 in 1995 to 700,000 in 1999 (IEA/SID,
2001). This was expected to rise to more than 1.5 million by 2002. The
problems faced by these children mimic a vicious circle going beyond
the challenges of household headship to child labour, lost education
opportunities, poverty and the likelihood of engaging in risky sexual
behaviour and crime (Steinberg et al, 2000).

In traditional family settings, AIDS orphans are cared for by members

of the extended family or the elderly. But the magnitude of the AIDS
problem has overwhelmed these structures in some parts of the country
(NASCOP, 1999). In 1994 an assessment by Plan International, Embu
branch (KIE, 1997), demonstrated that caring for orphans was already
a burden in many families. In general, deaths caused by AIDS have led
to dissolution of some households in parts of Western Kenya.

Population size

Using various demographic models, it has been predicted that AIDS

will slow the rate of population growth in Kenya by approximately one
percentage point by 2005. However, the population is expected to grow
by 1.3% by 2005. It should be noted that changes in the total fertility
rate will have a greater impact on the population growth rate than would
AIDS deaths (FHI/AIDSCAP, 1996). In a ‘no-AIDS’ scenario, the total
population would increase from 29 million in 1996 to 37 million by 2005.
With AIDS, the population is expected to reach 34 million by 2005.

27
HIV/AIDS in Kenya: A review of research and policy issues

Macroeconomic impact

HIV has greater economic impact than other endemic diseases such as
malaria because it primarily affects adults in their most productive years.
There are many channels through which AIDS affects macro indicators
such as GDP per capita, economic growth and level of investment, but
the main channel is through reduced labour productivity as a result of
absenteeism and loss of experienced workers, possible changes in labour
supply and demand, reduced industry profits, and falling domestic
savings. Further, since the wealthier owners of capital and the more
skilled and better educated subsets of the population have higher levels
of consumption and investment, command higher wages and are more
likely to be employers, AIDS is likely to have a great negative impact
on the Kenyan economy.

Using a ‘macro AIDS’11 model, the total cost of AIDS to the country as a
whole was projected to reach Ksh 4.1 billion in 2000 and Ksh 5.5 billion
by 2005 (FHI/AIDSCAP, 1996). This is expected to reduce Kenya’s GDP
by 14.5% within the next 10 years from levels expected if there were no
AIDS. At the same time, per capita income is projected to drop by 10%.
In other words, the economy will not only lose valuable members of its
workforce but will also witness diminshed resources available to the
survivors.

2.3 Government response to the AIDS epidemic

Recognising the seriousness of AIDS, the Government of Kenya has

come up with various policy initiatives to stem the scourge. The
Sessional Paper No. 4 of 1997 on AIDS in Kenya (GoK, 1997) recognised
the main response measures by the government as the establishment of

11
Adopted from that developed by Cuddington, J. and J. Huncock, 1994.

28
 HIV/AIDS in Kenya

the National AIDS Committee and the development of strategic plans

to deal with the scourge. Initially, the government established the AIDS
Programme Secretariat (APS) in 1985, which became the Kenya National
AIDS Control Programme in 1987. The 1987-1991 national medium term
plan emphasised the need for creation of awareness about AIDS, blood
safety, clinical management of AIDS and capacity building for the
management of AIDS control programmes at the national level. The
second medium term plan (1992-1996) stressed the need to involve all
sectors in HIV prevention in order to mobilise broader national response
against the epidemic (GoK, 1997). Also, Kenya’s Health Policy
Framework identified five goals in the fight against the spread of AIDS
(GoK, 1994):

• Prevention of HIV infection through information campaigns;

• Prevention of transmission through blood by effectively screening

blood;

• Prevention of perinatal transmission;

• Healthcare, counselling and social support for AIDS patients; and

• National coordination and mobilization of funds to cope with

the disease.

The government went further to recognise AIDS as a development issue.

This led to AIDS issues being incorporated into the Seventh National
Development Plan, the Fifth District Development Plans and other
succeeding policy documents. In 1999 the government declared AIDS
pandemic a national disaster. This allowed more government spending
and involvement in its fight, while creating opportunity for greater AIDS
related donor funding. The government’s response to the AIDS scourge
also involved resource mobilization to finance control and treatment of
AIDS-related opportunistic diseases. Apart from bilateral donors and
the World Bank, non-governmental organizations have been in the
forefront in providing financial support for the fight against HIV/AIDS.

29
HIV/AIDS in Kenya: A review of research and policy issues

The National Aids Control Council (NACC) in the Office of the President
now coordinates AIDS prevention and control activities in the country,
and is also the structure through which donor funds to fight AIDS are
channelled (mainly funding from the Global Fund). NACC’s National
Strategic Plan has five priorities for the prevention and control of HIV/
AIDS:

• Prevention and advocacy;

• Treatment, continuation of care and support;

• Mitigation of socioeconomic impacts;

• Monitoring, evaluation and research;

• Management and coordination.

The principle objective of the strategic plan is to stop the epidemic and
reduce its impact on the Kenyan society and economy by lowering HIV
prevalence in Kenya by 20 to 30% among people aged 15-24 years by
2005, increasing access to care and support for people infected and
affected by HIV/AIDS, and strengthening response capacity and
coordination at all levels (NACC, 2000).

The government has approved publication of the HIV/AIDS Prevention

and Control Bill, 2003, which seeks:

• To regulate education and information on AIDS;

• Safe clinical practices and procedures;

• Regulations on screening, testing and access to healthcare;

• Regulations on confidentiality such as on disclosure of

information and penalties for breach of confidentiality;

• Stemming discrimination in the workplace, schools and in

provision of services to HIV/AIDS people;

• Regulations on AIDS research.

30
 HIV/AIDS in Kenya

In 2001 the government published a National Condom Policy and

Strategy (GoK, 2001c) whose purpose was to improve access to
affordable quality condoms. This also includes effective assessment of
condom demand, social marketing of condoms and raising of funds to
finance condom supply and distribution. The targeted groups include
the youth, mothers attending antenatal and other mothers, commercial
sex workers, people living with HIV/AIDS and people who frequent
bars and drinking places.

With regard to gender aspects of the HIV/AIDS epidemic, the

government has published a strategic plan on mainstreaming gender
into the Kenya National HIV/AIDS Strategic Plan, 2000-2005. This takes
into account most concerns that are gender specific in the exposure to
AIDS and response mechanisms. This approach aims to empower
women and remove gender inequality at all levels of anti-AIDS
programmes (NACC, 2002).

Through these activities, the government has registered several

achievements, particularly in establishing systems to stem the steep rise
of the disease and in alleviating suffering of AIDS patients through
provision of medication.

Another important segment of AIDS intervention involves workplace

programmes and practices. The most common prevention activity is
education through distribution of reading materials, discussions on
AIDS, presentations and peer education programmes. Out of 16
companies surveyed in 1994 (FHI/AIDSCAP, 1996), 11 offered some
form of AIDS education and 94% of them provided condoms. Some
offered education, and STD testing and treatment, and others
counselling services. A few of them encouraged voluntary HIV testing.

31
HIV/AIDS in Kenya: A review of research and policy issues

2.4 Experience of other countries with HIV/AIDS

Uganda

Now considered one of the world’s earliest and best success stories in
overcoming HIV, Uganda has experienced substantial declines in HIV/
AIDS prevalence and incidence during the past decade. According to
Uganda’s Ministry of Health data, seroprevalence among 15-19-year-
old pregnant women, which is believed to reflect HIV incidence, fell
sharply from the early 1990s to around 1996 and has remained low
(Marum and Madraa 1999). Estimates show that national HIV
prevalence peaked at around 15% in 1991, and had fallen to 5% by 2001.
This dramatic decline is unique worldwide, and has been the subject of
curiosity since the mid 1990s and recent and intense scientific scrutiny.

Uganda’s HIV prevalence has been observed over time and across
different geographic and demographic populations. Uganda’s falling
HIV prevalence is likely to be associated mainly with a number of
behavioural changes. Changes in age of first sexual encounter, declining
casual and commercial sex trends, partner reduction and condom use
all appear to have played key roles in the decline. Although HIV
knowledge, risk perception and risk avoidance options can ultimately
lead to reduced HIV incidence, there is a complex set of epidemiological,
sociocultural, political and other factors that are likely to affect the course
of the epidemic in Uganda.

High-level political support with multi-sectoral participation set the

tone for the fight against HIV/AIDS. In 1986, Uganda’s new head of
state, Yoweri Museveni, responded to the epidemic with a proactive
commitment to prevention, emphasizing that fighting AIDS was a
patriotic duty requiring openness, communication and strong leadership
from the village level to the state house. In 1992, the multi-sectoral

32
 HIV/AIDS in Kenya

Uganda AIDS Commission (UAC) was created to coordinate and

monitor implementation of the national AIDS strategy. The UAC
prepared a National Operational Plan to guide implementing agencies,
sponsored task forces and encouraged the establishment of AIDS control
programmes in other ministries including Defence, Education, Gender
and Social Affairs. By year 2001 there were at least 700 governmental
and non-governmental agencies working on HIV/AIDS issues across
all districts in Uganda.

In 1986 Uganda established a national AIDS control programme (ACP),

which launched an aggressive public media campaign that included
print materials, radio, billboards and community mobilization for a
grassroots offensive against HIV. The ACP became the STD/AIDS
Control Programme in 1994 and has since trained thousands of
community-based AIDS counsellors, health educators, peer educators
and other types of specialists. Spreading the word involved not just
information and education but also emphasis on fundamental behaviour
change-based approach to communication and motivation.
Decentralization was actually a type of local empowerment process that
involved local allocation of resources — in itself a motivating force
(Green, 2002).

Since 1989 Ugandan teachers have been trained to integrate HIV

education and sexual behaviour change messages into curricula. At the
same time the government has been working to empower women and
the youth by allowing them more political voice. In addition, since 1988
respect and protection of the rights of those infected by HIV have been
inspired by a number of prominent Ugandan citizens and by public
events such as candlelight memorials and observance of the World AIDS
Day. The AIDS Support Organization (TASO) was established in 1987,
and has worked in advocacy against discrimination and stigmatization
while pioneering a community-based approach to the care of people
living with HIV/AIDS.

33
HIV/AIDS in Kenya: A review of research and policy issues

Religious leaders and faith-based organizations have been active in the

frontlines of the response to the epidemic. Mission hospitals were among
the first to develop AIDS care and support programmes in Uganda. In
1990, the Islamic Medical Association of Uganda piloted an AIDS
education project in rural Muslim communities that evolved into a larger
effort to train local religious leaders and lay community workers. The
Protestant Church of Uganda organized a workshop for bishops and
other religious leaders in 1991, and implemented an extensive AIDS
education project in many of its dioceses. The Catholic Church and
mission hospitals provided leadership in designing mobile home care
projects for AIDS victims and special programmes for AIDS widows
and orphans.

Uganda was the first country in Africa to launch confidential VCT

services. In 1990, the first AIDS Information Centre (AIC) for anonymous
VCT opened in Kampala. By 1993, AIC was active in four major urban
areas as more and more people became interested in knowing their
serostatus. AIC pioneered in providing same-day results using rapid
HIV tests, and the concept of ‘post test clubs’ to provide long-term
support for behaviour change to all those tested, regardless of their
serostatus.

In demographic health surveys, the ‘ever use’ of condoms as reported

by women increased from 1% in 1989 to 6% in 1995 and 16% in year
2000. For males, this was 16% in 1995 and 40% in year 2000. Modelling
suggests that very high levels of consistent condom use would be
necessary to achieve significant reductions of prevalence in a
generalized-level epidemic (Stover, 2002).

The most important determinant in the reduction of HIV incidence in

Uganda appears to be the decrease in sex partners and networks. In
general, Ugandans now have considerably fewer non-regular sex
partners across all ages. Population-level sexual behaviour, including

34
 HIV/AIDS in Kenya

the proportion of people reporting more than one partner in Kenya

(1998), Zambia (1996) and Malawi (1996), for example, appears
comparable to that reported in Uganda in 1988-1989.

The experience of Uganda suggests that a comprehensive behaviour

change-based strategy ideally involving high level political commitment
and with a diverse spectrum of community-based participation may
be the most effective prevention approach.

Thailand: the Phayao experience12

Phayao is a relatively small province in Thailand. Its main source of

income is agriculture, employing 44% of the workforce. The main crop
grown is rice. Low household incomes forced men to migrate in search
of employment leaving women behind. For women, one common option
was commercial sex work. HIV spread gradually since the late 1980s
turning Phayao Province the region with the highest reported AIDS
cases in Thailand. By 1994, AIDS, representing 11.3% of all deaths, had
become the leading cause of mortality in the province. The population
growth rate declined to almost zero.

The Phayao people countered HIV individually and collectively, mainly

through sexual behaviour change. Previously, the Phayao people
considered commercial sex as normal, and young men normally had
their first sex experience with a prostitute. This changed dramatically
when they came to understand that AIDS is a killer. For example, in
1990 there were 78 commercial sex establishments in Phayao with 449
commercial sex workers; these had dropped to 12 and 76, respectively,
by 1996. Also, condom use by male sex clients rose in most of the
brothels.

12
This section borrows from UNAIDS (2000).

35
HIV/AIDS in Kenya: A review of research and policy issues

Attitudes towards people with HIV/AIDS also changed from fear to

compassion. Adjustments were made to customs and lifestyles to
accommodate people living with HIV/AIDS. For example, infected
people were invited to participate in social events and to help in non
food-related chores. Twenty-four groups of people with HIV/AIDS were
established in Phayao with 1609 people. In addition, people started
dealing with the factors that made them vulnerable to HIV. Rather than
sending their daughters to town to make money, they started keeping
them in school with the help of the education ministry.

The government and NGOs also got involved. When HIV was first
detected in Thailand, the government appointed the sexually
transmitted diseases (STD) clinics as the main agencies to deal with the
new disease. They were put in charge of a prevention campaign focusing
on risk groups, mainly sex workers and their clients. In 1991, the
government created the National AIDS Prevention and Control
Committee (NAPCC) chaired by the prime minister. At the end of that
year, the NAPCC launched the ‘100% condom use campaign’ for
commercial sex establishments. The government has provided strong
media support for this campaign and also organizes condom
distribution.

From 1993 to 1996, the response to HIV/AIDS by Phayao institutions

steadily increased in intensity. In 1996, public and private institutions
in the province started 75 projects aimed at reducing HIV risk and
building capacity for dealing with HIV/AIDS. Many sectors were
involved in the projects, particularly at district level. Project funding
came mainly from the Ministry of Public Health. The projects benefited
many categories of people in Phayao.

36
3. Theoretical Issues and Empirical evidence

Studies on the impact of HIV/AIDS may be either at the macro or micro

level. At the macro level, HIV/AIDS affects growth through its impact
on macro variables such as interest rates and investment flows; physical,
human and social capital; and fiscal and revenue factors. Micro-level
economic effects are felt through reduction in household savings and
lowered labour incomes. Conversely, reduction in economic growth may
help fuel the pandemic. The direction of the impact of these factors on
the pandemic is largely an empirical issue yet to be resolved.

We start by looking at the micro impact aspects of the scourge together

with some empirical evidence on the health and agricultural sectors,
households and household economies, industry and population.

3.1 Micro- and meso-level effects

Industry and labour force

The main costs from the perspective of the firm may be classified as
direct or indirect costs. The direct costs include: 1) loss of hours of work
and specialised labour such as managers, entrepreneurs or those with
special modern technical skills, and 2) increased medical, insurance and
funeral costs (Steinberg et al, 2000). The indirect costs pertain to increased
cost emanating from: 1) absenteeism due to illness or funeral attendance,
2) provision of care to the ill, 3) lost skills, 4) training and recruitment,
5) reduced work performance, and 6) lower productivity.

Other factors influencing costs include the risk profile of the employees,
risk modification expenses of firms, and the degree to which work
processes are planned to accommodate people infected with HIV. For
most enterprises, costs associated with HIV/AIDS among employees

37
HIV/AIDS in Kenya: A review of research and policy issues

are unlikely to be devastating in any one year but can be substantial

over time and may affect international competitiveness in particular.
Illness or death of owners or key managers may be disastrous in some
firms. Small and medium-size enterprises reliant on local skilled people
are probably particularly susceptible.

Vulnerability to HIV/AIDS varies depending on factors such as type of

company and production processes. Labour-intensive firms may appear
to be at higher risk of lost production. The impact depends on the ease
with which employees can be substituted (Steinberg et al, 2000). High-
skill, labour-intensive industries such as those in education find it very
costly to retrain replacement staff, but low-skill industries such as
commercial cleaning can replace staff easily. Enterprises that offer
comprehensive benefits to employees are bound to suffer increased costs
as a result of AIDS.

The impact of AIDS extends to the marketing of goods and services.

Vulnerability of particular markets generally depends on the nature of
the good or service and the demographic and risk profiles of the
consumers. Certain markets, such as those providingg private healthcare
services actually benefit from the epidemic.

Various studies (Steinberg et al, 2000; Morris et al, 2000; FHI/AIDSCAP,

1996) have shown evidence of possible detrimental effects of AIDS on
industry. Steinberg et al (2000) observes that the impact on business
would vary depending on the type and the production processes. The
labour-intensive firms will lose if labour lost is not easily substitutable,
while capital-intensive firms will suffer if they lose specialised labourers.
Practically, the costs of the disease on the firm may be felt over the long
term, affecting both domestic and international competitiveness of the
firm. In a study in South Africa, Morris et al (2000) assessed the economic
impact of HIV infection in a cohort of rural agricultural workers from
the perspective of industry. The study found that HIV/AIDS seemed

38
 Theoretical issues and empirical evidence

to affect mostly people in the lower wage band, contradicting other

studies that show all workers are infected equally.

Impact on consumption spending

The main impact on consumption is likely to be in distributional terms.

This means that more income may go to treatment of AIDS-related
illness, while demand for luxury goods with high-income elasticities
will suffer. Therefore, the effect of AIDS depends on the extent to which
it depletes a household’s wealth and income base. There is general
agreement that the poor are hardest hit, and it is possible that AIDS
will increase the number of people below the poverty line. The sum
effect is lower consumer purchasing power, which will affect aggregate
demand. Other effects will be on banks and other lenders of credit. The
risk on loans is likely to be higher, which may have the effect of increases
in bank lending interest rates, leading to greater credit rationing and,
in effect, lower investment (Steinberg et al, 2000).

Agriculture

AIDS affects agriculture in as far as the poor and those mostly affected
depend on agriculture. It causes diversion of labour to the care
(homecare) of the sick or the total loss of farm labour, with the net effect
of lowering agricultural production and increasing food insecurity. In
general HIV/AIDS reduces food availability by lowering production;
reducing family labour, land and other resources allocated to agriculture;
and reducing livestock assets and implements. It also affects food
acreage and the stability and quality of food supplies as farmers shift
to less labour intensive production (Rugalema et al, 1998).

Illness and death can disrupt the farming cycle and reduce the ability
of households to produce and buy food. Sale of assets and weakened

39
HIV/AIDS in Kenya: A review of research and policy issues

social networks leave households vulnerable to production losses. With

lost labour, nutritional leafy crops and fruits may be replaced with
starchy root crops, such as cassava and maize, while livestock losses
reduce both manure and food supplies such as milk. AIDS also causes
farms to switch from export to food crops, with implications for export
earnings. Also, fertile land belonging to families hard hit by the epidemic
may remain idle due to labour shortage. In the late 1980s in Uganda
people reduced their work on coffee, which required pruning and
marketing, in favour of their staple bananas. They then cut down labour
on bananas and vegetables and concentrated on easily cultivated, easily
stored and starchy cassava. This shift to lower quality foods is associated
with chronic food insecurity, high levels of protein-energy malnutrition
and deficiencies in micronutrients such as iron, zinc and vitamins. This
further undermines the nutritional requirements of AIDS sufferers to
fight opportunistic infections and prolong survival (Loenwenson and
Whiteside, 2001).

Labour losses due to AIDS deaths have led to significant productivity

losses in much of Africa. In Ethiopia, for example, labour losses due to
AIDS reduced time spent on agriculture from 33.6 hours to 11.6 hours
per week. In Zimbabwe AIDS led to declines in maize outputs of 61%;
for death from other causes this was 45% (Loenwenson and Whiteside,
2001).

Impact on households and household economies

The initial impacts of HIV/AIDS on the household pertain to loss of

insurance and medical benefits, increases in medical and other costs,
and loss of the ability to work once one develops full-blown AIDS. This
means that ailing family members, who are usually at their most
productive age, become more dependent on the elderly or young
uninfected members of the household.

40
 Theoretical issues and empirical evidence

Death in the household from AIDS can have profound implications for
resource allocation, production, consumption, savings, investment and
the well-being of survivors (Steinberg et al, 2000; Arndt and Lewis, 2000).
AIDS comes with disruption of normal social relationships within the
family. This refers to the breakdown of usual relationships and support
systems, especially when the head or key members of the household
die. The threat of poverty becomes real as medical bills lead to a
depletion of income and savings. The survivors may be left with little
income to spend or wealth to inherit.

The costs to a household with an AIDS patient can be divided into three
components: direct costs associated with medical expenses, indirect costs
of foregone earnings, and indirect costs to other households who
contribute to funeral expenses and care for the orphaned children.

Studies in Côte d’Ivoire, Burundi and Haiti (UNAIDS, 1995) identified

socioeconomic impacts of AIDS that included cessation of paid
employment, increased borrowing, sale of possessions, low levels of
seeking care even when care was available, and ostracism and
discrimination of those affected by the disease. In Tanzania, India and
Thailand, the immediate impacts of AIDS were a noticeable drop in
income, an increase in debt and mortgaging, withdrawal of children
from school, early entry of children into the labour market, continued
presence of older household members in the labour market, and effects
on the form and patterns of employment of care providers.

Contrary to common expectation, a study by Bertozzi and Aggleton

(1995) found that based on aggregate data, there was no direct link
between HIV/AIDS and poverty. But Arndt and Lewis (2000) found in
Botswana that: 1) a quarter of all households would lose income for
over a decade as a result of death from AIDS of an income-earning
member, 2) there would be tremendous increase in the number of
families without income, and 3) the epidemic would cause a decrease

41
HIV/AIDS in Kenya: A review of research and policy issues

of between 8 and 10% in household per capita income over the next 10
years. Though income inequality was unlikely to be affected by AIDS,
the percentage of people living in poverty is likely to increase by between
4-6% as a result of AIDS. We should, however, note that the number of
poor people would in fact be much higher if the poverty line was
redefined to take into account the preservation of life itself as a basic
need. This would require us to factor in the total cost of HIV treatment.

Another factor in the impact of AIDS is the ability of households to

cope with the disease, the key channels available for insurance from
risk and sources of assistance or credit. Research has shown that
vulnerability to shocks varies among households, and this may
significantly affect the path of development and distribution of welfare
(Lundberg et al, 2000). Households respond in various ways to shocks
such as an AIDS shock. Wealthier households seem to be better insured
than the poorer households against shocks, owing to their access to
credit and richness in social capital, i.e. they have a larger, broader and
presumably wealthier network of friends and relatives on whom they
depend in times of crisis. Poor households normally tend to rely on
private transfers rather than on credit.

A bereaved family may cope with economic shocks caused by death by

1) seeking financial assistance from friends and relatives, or a formal
government or NGO agency, 2) changing the mix of crops grown on
the family plot, 3) altering members’ time allocation between labour
market participation and work at home or school, 4) selling assets, 5)
recruiting or shedding household members, or 6) reducing consumption
of some or all household members. Research shows (Lundberg et al,
2000) that households that suffer an AIDS death receive more net private
transfers and assistance from NGOs and the government than do
households that have not suffered such a calamity; but the difference is
not significant (Lundberg et al, 2000).

42
 Theoretical issues and empirical evidence

Private transfers provide by far the greatest assistance, but not all
households rely on them (Lundberg et al, 2000). The dependence of some
households on assistance from public institutions and NGOs could
crowd out private assistance. Estimates of this crowding-out effect range
from negative one-for-one to positive, where public assistance actually
stimulates private transfers. In the long run, the ability of a household
to cope with an AIDS shock diminishes as the ‘autocorrelation’ in shocks
rises or as the shock is repeated over time. This indicates that households
with repeated AIDS attacks have the least ability to cope.13

Impact on health sector

The AIDS epidemic has increased the demand for public health about
seven times in most African countries, doubling bed occupancy rates
and crowding out the needs for other health conditions. The annual
direct costs for AIDS (excluding antiretroviral therapy) is estimated at
about $30 per capita, while most African countries spend less than $10.
Inadequate resources for AIDS have put health systems under pressure,
eroding the quality of care (Loenwenson and Whiteside, 2001). The
health sector workers may obtain infection in the line of duty as they
come into contact with infected patients. They are also stressed up in
cases where health centres are poorly funded due to the magnitude of
the epidemic.

But there is also a possibility that non-AIDS related health spending

might be declining as a result of demographic changes (Steinberg et al,
2000). For example, early in the epidemic, a decrease in the number of
women of child bearing age will lead to a decrease in pregnancy and
birth related expenditure, and also due to the number of old people

13
In a well-established epidemic, households may learn from experience or
from other affected households to insure from frequent AIDS deaths.

43
HIV/AIDS in Kenya: A review of research and policy issues

entering old age. Therefore, as the HIV related expenditure (on

antiretroviral, bed occupancy, etc) rises, the non-HIV related expenditure
would be declining.

Education system

AIDS affects both the availability and use of schooling in terms of the
number of children enrolled in school and the number of teachers and
education officials (Loenwenson and Whiteside 2001). School-age
children are forced to take care of their sick parents or relatives, and
most cannot afford fees, lack parental care and face hunger and
emotional stress. The situation is particularly bad for AIDS orphans
and HIV-positive children.

Studies in Guatemala show that more than a third of children orphaned

by HIV/AIDS drop out of school (Loenwenson and Whiteside, 2001).
The World Bank estimates that by year 2020, Tanzania will have lost
some 27,000 teachers to the disease, and their replacement will cost USD
37.8 million (World Bank 1992). UNICEF estimates that 860,000 children
in Africa have already lost their teachers to HIV/AIDS (Loenwenson
and Whiteside, 2001). In Botswana, death rates among primary school
teachers rose from 0.7 per 1000 in 1994 to 7.1 in 1999 (Loenwenson and
Whiteside, 2001). At the same time, the total population paying for
education will increase more slowly than in a no-AIDS scenario.

3.2 Macro-level effects

Although some studies (Nicholls et al, 2000; Bonnel, 2000; Arndt and
Lewis, 2000) posit that the economic cost of AIDS at the macro level
will be great, more conservative research shows that the actual impact
of HIV will be small, not catastrophic (World Bank, 1997)14. However,

14
The World Bank (1997) report on confronting AIDS contains a sub-section
entitled “AIDS has little net macroeconomic impact”.

44
 Theoretical issues and empirical evidence

there is general agreement that the socio-economic cost will be big. This
is in terms of loss of life, family dissolution and crime increase, among
other costs. The pandemic has been seen to have effects on macro
variables such as physical, human and social capital.

HIV/AIDS also affects the macro outcomes through its effect on fiscal
deficits (Bonnel 2000; Arndt and Lewis, 2000). This is through increasing
government health expenditure on its own employees and on the public
health budget. All these worsen the fiscal deficit of a country and reduce
domestic savings. This will most often be the case in a developing
country such as Kenya, which cannot offset such deficits through cutting
expenditure or raising taxes. Note also that as production falls, or as
shifts occur in the private sector due to HIV/AIDS, the government
revenue base, especially the VAT and corporate taxes, and trade taxes
will also be affected negatively. AIDS forces households to deplete their
assets and savings, which also affects government revenue collected
through income taxes.

Whether savings rise or fall in affected households is ambiguous.

Evidence shows that savings fall in the face of the disease (FHI/
AIDSCAP, 1996; Bonnel, 2000; Arndt and Lewis, 2000). Household
savings may fall due to death of adult family members and depletion
of assets and savings. However, savings could increase in a well-
established AIDS epidemic, where households are forced to save more
in anticipation of increased spending when AIDS infection strikes, or if
HIV/AIDS leads to rise in per capita income. This may only happen
where skilled labour is in short supply.

Generally, evidence (FHI/AIDSCAP, 1996; Arndt and Lewis, 2000;

Nicholls et al, 2000; Bonnel, 2000) shows that an HIV/AIDS epidemic
would cause a fall in savings. The combined effect of decreased domestic
savings is lower investment, assuming that there is not enough external
financing to offset the fiscal deficit. Initially HIV/AIDS has negligible

45
HIV/AIDS in Kenya: A review of research and policy issues

effect on physical capital, but over time, as the epidemic takes its toll,
the reduction in domestic savings affects investment. Physical capital
is also affected through worsening fiscal deficit resulting from increased
health expenditure on AIDS and related illnesses, pension payments
for AIDS-related deaths of civil servants, and training of new employees.
This affects the government’s ability to invest.

The impact of HIV/AIDS on human capital is stronger in the short term

than in the long term. The initial effect of HIV/AIDS is to destroy human
capital, affecting both the high and low-income groups. In the long term,
the high income groups can lower their susceptibility to AIDS through
larger investment on prevention of the scourge. The initial decline in
labour supply associated with AIDS infection leads to increase in return
to skilled labour and in wages for skilled labour (if there is no surplus
labour). This may not work in countries like Kenya with excess labour).
To the extent that the wages of skilled labour rise relative to those of
unskilled labour, the retun to schooling would increase. This would be
especially relevant for the youth who are not yet HIV positive. So far,
available evidence (Bonnel, 2000) suggests that formal education is
adversely affected by HIV epidemic.

Social capital has emerged as an important determinant of growth. For

example, Bonnel (2000) notes that trust and civic cooperation matter in
economic growth. HIV affects the social structure of local communities
and erodes existing social networks and traditional support systems.
Other effects are increased insecurity, especially in terms of increased
crime resulting from the increased numbers of orphans with no adult
guardians or income sources (Bonnel, 2000).

HIV/AIDS generally worsens poverty at the macro level. A study in

Botswana (Greener et al, 2000), for example, showed that about 26% of
poor households surveyed would lose their income as a result of death
of an income earner, while small households would be wiped out over

46
 Theoretical issues and empirical evidence

a ten-year period because all of the household members are infected

(Greener at al, 2000). Generally, poverty levels would increase and per
capita incomes would decline by 10% over the next 10 years. Though it
is widely agreed that HIV/AIDS would generate a fall in output, the
impact on per capita incomes is indeterminate ex ante (Bonnel, 2000).
Early analysis of impact of HIV/AIDS on per capita incomes in
Botswana (Greener et al, 2000) show that the epidemic could increase
per capita income as population declined. However, evidence generally
shows that increased public and private AIDS-related health
expenditure would lower the stock of capital in many African countries
through a fall in investment, caused by a fall in domestic savings. This
in effect lowers per capita incomes. Botswana, whose per capita incomes
may increase due to the big national savings and dependence on
diamonds for export, would be an exception in this case.

Cross-country studies confirm the negative effects of HIV/AIDS on

economic GDP and average growth rate per capita (Bonnel, 2000). It
has been found that had the HIV prevalence not reached 8.6% in 1999,
Africa’s income per capita would have grown at 1.1% per year. In the
case of sub-Saharan Africa, for a nation with a prevalence rate of 20%,
the rate of GDP growth over 20 years would be 67% less than would
have otherwise (Bonnel, 2000). Nicholls et al (2000) show that in Trinidad,
Tobago and Jamaica, a sustained increase of 20% in the number of people
infected between 1997 and 2005 would lead to contractions in GDP of
between 4.2 and 6.4%, and AIDS may slow down capital accumulation,
assuming that investment is financed from domestic savings, health
spending is drawn from domestic savings and there are no external
grants. Normally, the effects of AIDS are likely to be felt over time in
terms of reduced rate of accumulation of knowledge (in terms of reduced
total factor productivity) and falling rate of accumulation of capital
through erosion of human capital and switching of domestic savings to

47
HIV/AIDS in Kenya: A review of research and policy issues

finance current health spending (Nicholls et al, 2000; Arndt and Lewis,
2000).

Most of the studies have also focused on the effect of AIDS on certain
sectors, the fiscal position and demographic outcomes. Using a
computable general equilibrium model and incorporating the impact
of AIDS, comparison was made of a hypothetical ‘no-AIDS scenario’ to
a more likely ‘AIDS scenario’, and the decomposed model was used to
determine the factors that contributed to the different growth levels.
Using this kind of modelling, real GDP was about 17% below the level
attained in the no-AIDS scenario. However, the differential or the size
of GDP as a measure of economic welfare may not be satisfactory,
because it may not be an adequate measure of welfare in the context of
an epidemic that lowers the average life expectancy by 20 years (Arndt
and Lewis, 2000).

Many of the predictions on the impact of AIDS on the economy differ

greatly in methodologies chosen and assumptions used. Many aspects
of the direction the impact will possibly take are not clear, but there is
agreement that AIDS bears heavily on the economy in the short and
long term.

48
4. AIDS Research Methodologies in Economics

Macro growth models

Macro modelling of the impact of AIDS on the economy was pioneered

by Cuddington and Hunckock (FHI/AIDSCAP, 1996). The macro model
they developed has been used to assess the impact of HIV/AIDS on
the economies of Tanzania, Zambia and Malawi, a modified version in
Kenya (FHI/AIDSCAP, 1996). The model makes projections of some
economic indicators based on time series data of economic performance
with AIDS epidemic. The projections made, based on the assumption
of the presence of AIDS, are compared with those made under the
assumption of its absence. Past information on the economic
performance before the AIDS epidemic is required to make projections.

The series data used to build an AIDS scenario for Kenya spanned nine
years, from 1985 to 1993, when interventions on AIDS were still
rudimentary and prevalence rates low, and would not be adequate to
justify projections to 2010. This model is good for predictions used to
bring attention to possible effects of the AIDS scourge but would be
unrealistic in attributing real economic costs, since the model is based
on many strong assumptions.

The impact of HIV/AIDS can also be estimated using a growth model

with the prevalence rate of HIV as an explanatory variable. Though a
biased estimate of the impact of HIV on the economy, GDP per capita
over time is usually used. The per capita GDP neglects the welfare of
those who have died of the scourge. In modelling the impact of AIDS
on the economy, it is important to take into account identification
problems. It is unclear whether HIV affects growth or growth affects
HIV. A plausible solution to the identification problem is to use various
equations to address various interactions. Bonnel (2000) used three

49
HIV/AIDS in Kenya: A review of research and policy issues

equations in a cross-country study. One equation expressed growth as

a function of macroeconomic policy ratings, institutional variables and
other determinants of growth. Basically, the average rate of growth of
GDP per capita was expressed as a function of initial per capita income
of a country, level of infrastructure (per capita number of telephones)
and index for legal framework, macro policy ratings for the country,
human capital and other variables used for their significance in the
model (such as malaria prevalence).

The second model expressed policy ratings as a function of initial

macroeconomic ratings, natural logarithms of initial per capita income,
infrastructure, and HIV prevalence and its square. In the third equation,
the natural logarithm of prevalence rate of HIV/AIDS was expressed
as a function of the main determinants of HIV/AIDS. These include
infrastructure, average growth of GDP per capita, extent to which
different (ethnic) languages are used, share of female labour in industry
in 1990 (representing gender), share of receipts in exports in 1990
(representing migration), share of the Muslim population, malaria
morbidity in 1990, number of years since the first HIV/AIDS case was
reported, and secondary school enrolment in 1990. The OLS estimation
procedure was used together with three stage least squares, which takes
care of endogeneity problems (Bonnel, 2000).

Interesting as this approach may seem, data for a single country may
limit to some extent the usage of endogenous growth models. However,
assuming that time series data on HIV prevalence rates and per capita
GDP were available, Bonnel (2000) defines the following specification:

G = f(H,H2, HC, D, Z, LAW, M).................................................1

Where: G is the GDP per capita in year t, H is HIV prevalence in year t,

H2 is the square of HIV prevalence for year t, HC is the human capital
stock in year t, LAW is a measure of the rule of law (proxied by the
corruption index), M stands for sound fiscal and monetary policies (may

50
 AIDS research methodologies in economics

be proxied by the budget deficits or the rate of inflation), D is malaria

prevalence, and Z are variables tested for their significance (these would
include an index for the development of the financial system and
dummies for civil unrest, among other variables).

Since HIV prevalence is suspected to be endogenous in the model, we

may estimate it separately before using it as an explanatory variable in
the model above. Covariates entering the HIV prevalence model were
infrastructure per capita each year (e.g. number of phones per capita),
a gender equality index, secondary school enrolment in year t, and
number of years since the first case of HIV was reported for each year.

Experimental approaches: impact of HIV/AIDS on poverty and

inequality

The impact of HIV on poverty levels may be investigated using an

experimental survey where a random number, say 0 or 1, based on the
prevalence rates, is assigned to an experimental sample (Greener et al,
2000). The simulated infection pattern should resemble the age, sex and
geographical distribution of HIV/AIDS in the target population.
Normally, data from sentinel surveillance would do. Using a
predetermined poverty line and a hypothetical period within which an
AIDS patient would die, the monetary costs of AIDS can be computed.
This methodology also allows analysis of the expenditure and income
effects. The poverty line can be adjusted to reflect additional medical
and funeral expenses, therefore raising the likelihood that a family with
an HIV/AIDS patient would be poor, and at the same time taking into
account possible benefits across the whole sample of higher employment
and higher wages for skilled workers with increased AIDS deaths
(Greener et al, 2000). This method seems attractive in estimating the
HIV/AIDS effects on institutions, specific districts or regions.

51
HIV/AIDS in Kenya: A review of research and policy issues

Economy-wide CGE models

A Computable General Equilibrium (CGE) model is based on a

consistent and balanced social accounting matrix (SAM) that requires
that key behavioural and accounting constraints and balance of
payments equilibrium be maintained. These models are rich in sectoral
and income distributional data. They are widely used to evaluate trade
and fiscal policies affecting taxation and expenditure, since these have
varying impacts on different productive sectors and income classes.
They are underutilised even though they are best suited to analyse the
economy, and are specially adapted to the economic analysis of the
impact of AIDS. A CGE model can be used to simulate the impact of
HIV/AIDS on all sectors of the economy. Arndt and Lewis (2000) have
used a CGE model. The model assumed that HIV/AIDS spending is
deficit-financed and net foreign savings are exogenously determined,
and allowed for varying fixed and exchange rates to achieve external
balance (Arndt and Lewis, 2000).

Production function approach

HIV/AIDS, just like malaria, reduces the amount of labour available to

industry, therefore reducing output. AIDS mortality and morbidity affect
production in firms owing to its effect on quantity and productivity of
labour. For skilled labour or workers with special skills, the loss is
significant because the cost and time involved in replacing such labour
are high. The effect of HIV on household production, for instance, can
be investigated using a production function.15 The effect of HIV/AIDS
on production can be captured using a specification with HIV prevalence
as an explanatory variable. The nature of the production function,
whether Cobb Douglas or flexible type, should depend on our objective

15
See the draft paper on “Impact of malaria on crop production in Kenya”
(2003), KIPPRA.

52
 AIDS research methodologies in economics

and data type (Berndt and Christensen, 1974; Douglas, 1976; Denny
and Fuss, 1977).

Cost of illness approach

The cost of illness approach (COI) incorporates the direct and indirect
costs of illness with direct costs consisting mainly of medical treatment
costs, and indirect costs consisting of the value of productive time lost
due to illness. Following Malaney (2000), the COI approach adapted
for AIDS cost analysis would involve summing private medical costs,
non-private medical costs or government spending on anti-AIDS
treatment, labour loss due to mortality and morbidity, risk-related
behaviour change costs, investment losses, and non-economic personal
burden, which includes a costing of intangible losses occasioned by pain
or suffering and other inconveniences due to AIDS.

The COI may also incorporate the value individuals attach to labour
and the way households form their expectations about future illness
costs (Cropper et al, 1998). Households may form their expectations by
assuming that their future illness costs will exactly equal previous year’s
costs. This means that COI = COI*. The alternative involves their looking
beyond their experiences and calculating their chances of contracting
HIV/AIDS using both the relative frequency of the disease in the
population and the average COI. Consequently, the chance of an
individual contracting the disease is the relative frequency of the disease
in the sample. Note, however, that due to the stigma attached to AIDS
the innate chance of infection could be overstated.

Willingness to pay approach

The willingness to pay (WTP) approach entails eliciting from households

or individuals the amount they would like to pay for hypothetical

53
HIV/AIDS in Kenya: A review of research and policy issues

prevention or eradication of HIV/AIDS. One practical way of doing

this is to carry out a survey on amounts individuals would pay for a
hypothetical and effective AIDS vaccine.16 Such a study would need to
be carried out in a place with high AIDS prevalence. A few caveats may
suffice in an AIDS study as opposed to a study on malaria, for example.
One, AIDS is stigmatised and, two, people do not like enquiries on their
sexual behaviour, which they consider private. Therefore, the values
that may be attached to preventive methods such as special condoms
or a vaccine may be either highly understated or overstated, depending
on individual opinion.

Another conceptual issue points to the possible aggregation of WTP

across the population. This will not pass, owing to the shortcomings
mentioned above and due to the reason that we cannot sum individual
welfare functions to obtain a social welfare function. Further, values
attached to prevention programmes (such as vaccines) may differ from
values households would attach to total eradication of AIDS. Finally,
implementation of COI and WTP approaches requires expensive survey
data. Survey studies may also incorporate situational and impact
assessment in certain sectors or among certain groups of people affected
by AIDS.

Micro studies: Institutions and AIDS patients

We have extensively reviewed various constraints and problems

industries and institutions such as schools, colleges and hospitals face
due to HIV/AIDS. Worker absenteeism, medical and insurance costs,
loss of experienced staff and funeral costs are some of the costs that
firms deal with. This clearly means that most of these firms must be

16
This kind of study has been carried out to quantify the burden of malaria
(Cropper et al, 1998). This study envisaged a hypothetical malaria vaccine and
a bed net scenario in prevention of malaria in the sample survey.

54
 AIDS research methodologies in economics

taking direct or indirect steps to curb the impact of this disease. Key
among such interventions are specific programmes for worker
sensitization, provision of preventive materials to workers, among other
mechanisms. Therefore, there is need to study what actually the top
managers of various firms do about the scourge. Their opinions and
intervention mechanisms may help in funding further AIDS control and
prevention measures. A study may therefore be necessary to collect the
views of management on the scourge and document different
programmes that are in place, sources of funding, and the programmes’
effectiveness and problems. This could be supplemented directly by,
say, a workers’ questionnaire module to elicit their opinions about the
disease and their role in reducing its spread.

Such a study may not be complete without a module on those living

with HIV/AIDS, who most likely are found in various homes or
hospitals. This may help in understanding various problems these
people undergo and in identifying ways of reducing the burden of the
disease. Another important area of research is the economic burden from
children orphaned by the disease and the economic and social cost of
family break-up.

55
HIV/AIDS in Kenya: A review of research and policy issues

5. A Case for Further Research

From the research review we can surmise that studies on the economic
impact of HIV/AIDS have focused on macroeconomic, household and
industry levels. In Africa, macroeconomic impact studies have been
carried out in Tanzania, Cameroon, Zambia and Kenya among other
countries; household studies in Cote d’Ivoire, Uganda and Ethiopia
(Arndt and Lewis, 2000), and industrial studies in Tanzania, Malawi,
Zimbabwe, Zambia and Kenya. Though these studies are breakthroughs
in AIDS impact research, their use of incomplete epidemiological data
heavily affects their reliability in policy analysis.

Despite the large scale of the epidemic, data are scanty on the impact of
the scourge at the individual, community and national level (Steinberg
et al. 2000). This is associated with inherent difficulties in collecting
complete and unbiased data owing to the nature of AIDS and the fact
that anonymous sentinel surveillance data are used. Further, predictions
regarding the impact of the epidemic are mainly based on the
mathematical simulation models (such as the Doyle simulation model)
that use sentinel surveillance data. These data are largely biased, since
they exclude children and the elderly from the study, and the fact that
they are collected from antenatal clinics indicates that they seldom
capture the lower risks of men. Since women are at higher risk of getting
infected with HIV than men, using antenatal clinics data implies that
we are assuming that both men and women are at the same level of risk
of getting HIV.

Owing to the nature of data available for AIDS research, most studies
of the epidemic are rather speculative. Besides, precedent exists in health
economics of studies of an epidemic of this magnitude. Nevertheless,
given the inertia of the problem to subvert economic gains, sometimes
hasty, theoretical or predictive studies are necessary insofar as far as
they broaden our understanding of the problem and can help us plan.

56
 A case for further research

Some salient factors point to the need for economic research on AIDS in
Kenya. For instance, what are the main channels through which HIV/
AIDS impacts on the economy in Kenya? How do we use the poor or
non-representative data to make robust predictions of the economic and
social cost or impact of HIV/AIDS on the economy? Given the prevailing
HIV/AIDS situation and in the light of the current economic slowdown,
what are the policy options for the government and the private sector
to arrest the scourge and absorb the predictable and inevitable impacts
of the disease? How will this be done in the light of unforeseen economic
shocks or national or global crisis? What is the best policy mix, or what
combination of programmes achieves the best outcomes in terms of
minimized AIDS costs and reduced infections?

It is also important to understand the role of intergenerational aspects

of the AIDS impact and stigmatization in the spread of the disease.
Stigmatization may help fuel the scourge in addition to affecting the
welfare of those affected. Therefore, it is could be worthwhile to estimate
the economic benefit of eliminating stigmatization on reducing new
infections, for example.

Intergenerational aspects refer to the impact of AIDS on a household

and a community in the long term. The aftermath of AIDS will
undoubtedly lead to increased number of orphaned children and
poverty dynasties with depletion of household savings and death of
key income earners. More research is needed focusing on the true loss
to a household from an AIDS death at the present and across future
generations. This should also include enquiries into household
responses to the disease and how the scourge affects household
consumption choices and saving preferences in both the short and the
long term. The research should also note that each new infection places
certain demand on current and future healthcare and social welfare

57
HIV/AIDS in Kenya: A review of research and policy issues

resources based on current estimates for the cost of treating an AIDS

patient and cost of economic aid to such individuals, among other costs.

Studies on the impact of AIDS at the workplace and on firms need to

take into consideration infection by skill or income class. This can help
in implementing workplace interventions. Investigation of AIDS impact
on the upper income classes, owners of capital and high-ranking
managers, would be extremely valuable, owing to the importance of
this class in investment and economic growth. Also, AIDS studies in
Kenya may need to take into account the welfare impact on the
individual and households. Nicholls et al (2000) for example suggest
that further research should incorporate the welfare impact of the
epidemic through a social welfare function.

Research on the macro impact of HIV/AIDS in Kenya is scanty. There

is need, therefore, to assess the effect of the scourge on macro variables
such as growth and interest rates, employment levels, physical and
human capital and savings. These should be seen as different from
sectoral or micro studies that focus on a particular industry such as
agriculture, or on effect on households, individuals or AIDS patients.
We note, however, that sectoral effects feed into macro impacts, which
means that analysis always should be twofold: 1) analysis of the impact
of AIDS at the ‘local’ level, for example at the household, community,
individual, single-business, a particular industry, or an industry-group
level, and 2) analysis of the sum of the individual impact on macro
variables or indicators at the national level.

The literature reviewed in this study provides various approaches

employed to model the impact of AIDS at the macro level. We have
appraised the production function and WTP approaches to cost
estimation, alongside economy-wide models and the growth equation
approach. Sectoral and workplace studies serve to open up information

58
 A case for further research

on the situation on the ground and inform on the best interventions

necessary to curb the spread of AIDS. The impact of various anti-AIDS
interventions also needs to be measured and justified.

59
HIV/AIDS in Kenya: A review of research and policy issues

6 Priorities and Policy Issues

In the shadow of the sustained AIDS problem in Kenya, the very

important but two-pronged policy question is how the government can
minimize cost of AIDS while reducing new infections. Programmes
designed to reduce economic costs of AIDS cannot be de-linked from
those that help reduce new infections. Both kinds are interlinked and
reinforce each other. First, costs of AIDS may be reduced by taking steps
to reduce labour losses (destruction of human capital) and therefore
increase productivity. Since the greatest economic impact of AIDS is
through destruction of human capital, efforts to offer life-prolonging
treatment will reduce this loss and minimize total loss to the economy.
Therefore, the economy can incur the costs of providing life-prolonging
medication since the benefits in terms of human capital saved and other
benefits are greater than the cost of the intervention. Other interventions
targeted at cutting economic losses may include economic and food
aid for the infected or affected, social support and retention of infected
people in their jobs.

Efforts designed to reduce new infections mainly through prevention,

awareness campaigns and VCT services should be intensified. Taking
the test is a very important outcome of awareness and prevention
campaigns, therefore the significant place that these campaigns (should)
occupy in the war against AIDS. VCT services have gained prominence
worldwide as a keystone for all anti-AIDS activities. Great benefits
accrue from early diagnosis of HIV status, since those affected can
promptly start taking life-prolonging ARVs and adopting other steps
to improve their lives. For the government, VCT data are invaluable;
they form a backbone for planning and spending on healthcare, and
economic and social support for those affected. Most targeted AIDS
interventions should be centred on the outcomes of correct data on

60
 Priorities and Policy Issues

diagnosis of individuals in Kenya. Even awareness campaigns should

be anchored on data from VCT services.

Policies and interventions that gravitate around and are informed by

VCT services, however, may not be sustainable owing to the low number
of people willing to be tested for HIV/AIDS. Therefore, there is need to
promote testing through creating appropriate incentive mechanisms.
This means incentives must be put in place to encourage obstinate
people to know their HIV status. Primarily, the choice to take the test
will depend on implications of the results of the test and incentives
available to the individual.

Awareness campaigns, though indispensable, are likely to be widely

accepted if they can show proof that VCT centres are accessible and
that there is a clear support mechanism regardless of the test outcome.
Adequate numbers of VCT centres and proper individual follow-up
mechanisms after the test must be in place, providing services such as
specific treatment regimes, care, social support assurances and legal
protection for those testing positive, and means of and opportunities
for prevention, random follow-up and aid to offset risky behaviour for
those testing negative.

Although VCTs are the loci for HIV/AIDS interventions and campaigns,
chances to obtain desired behavioural reaction, that is, deliberate choice
to take a test, greatly depend on the incentives available. The incentives
can broadly be categorized as economic, social support, treatment
availability, prevention, and proper laws to protect HIV-positive people
and test takers (HIV/AIDS Prevention and Control Bill, 2003).

Medical care should be top on the menu of incentives. The government

should ensure that those who test positive receive appropriate life-
prolonging medication and are treated for opportunistic infections.
Economic support may include food and nutrition aid, job security and
support to educate orphans. There should also be a way for ensuring

61
HIV/AIDS in Kenya: A review of research and policy issues

family or community support, including support for children and

protection from domestic violence. Further, HIV-positive persons can
be provided with protection to prevent their re-infection or infection of
others. All these interventions could be coordinated by an integrated
response centre that would hold databases on all positive tests, real
and simulated, according to regions and that would allocate resources
accordingly. Such a centre could be replicated at any administrative
level including the village. The data may also be used for future
economic impact programming. There should be clear and tangible
mechanisms to help HIV-negative people to remain that way. This could
include provision of prevention aids, counselling and a deliberate
attempt to offset aspects that put them at risk. A risk factor such as
unemployment or prostitution may be offset by employment creation,
which itself is entrenched in the government’s Economic Recovery
Strategy paper.

These proposals are a logical organization of ideas and current practices

that governments and NGOs espouse but that are so fragmented that a
vibrant intervention system would not be discernible. Further, owing
to the nature of the AIDS pandemic–that the epidemic is stigmatised
and that most interventions tend to shake strongly held cultural and
traditional values–programmes against AIDS must be well thought out.
They must involve all stakeholders and be credible, until a point is
reached where there is a sustained momentum nationwide to fight all
aspects of AIDS.

Issues of governance and management of programmes need greater

attention, particularly with regard to use of funds. Monitoring and
evaluation systems and analysis of cost-effectiveness of anti-AIDS
projects should be mainstreamed into national NGO or government
intervention drives.

62
 Priorities and Policy Issues

People who show compassion for those affected by the scourge–not

opportunists and profiteers who infiltrate this humane war against
AIDS–may also be allowed to run Anti-AIDS projects. There should be
a system of vetting those who run such programmes, since their
credibility influences the perception of the public, and especially those
affected, of anti-AIDS messages and whether they support the projects,
drives or strategies to fight the scourge.

The key inputs to what could be called a national anti-AIDS programme

are:

• Data on early-diagnosis and a follow-up programmes, to allow

for animation of case-specific interventions, with voluntary
counselling and testing (VCT) forming the backbone for medical,
economic and social support regimes;

• Collection of incidence and general prevalence data to serve as

impact indicators and for impact analysis and long-term planning;

• Favourable and sustainable linkage of all awareness campaigns,

VCT drives and follow-up strategies, treatment regimes and social
and economic support programmes to the PRSP and the
budgeting process (MTEF), with greater commitment of
government allocations to these programmes rather than
dependency on donor funding.

63
HIV/AIDS in Kenya: A review of research and policy issues

References
Arndt C. and J.D. Lewis (2000). The macro implications of HIV/AIDS
in South Africa: a preliminary assessment. The South African
Journal of Economics, 68(5):856-887.
Berndt E.R. and L.R. Christensen (1974). Testing for the existence of a
consistent aggregate index of labour inputs. The American
Economic Review, 64(3): 391-404.
Bertozzi S.M. and P. Aggleton (1995). Report from a consultation on the
socio-economic impact of HIV/AIDS on households, Chiang Mai,
Thailand, 22-24 September 1995.
Bonnel R. (2000). HIV/AIDS and economic growth: a global perspective.
The South African Journal of Economics, 68(5): 820-855.
Boseley S. (2003). Africa’s deadliest battle: saving grace. Africa Today, 9
(5): 19-36.
Cropper M., H. Mitiku, J. A. Lampietti, C. Poulos, and D. Whittington
(1998). The value of preventing malaria in Tembien, Ethiopia. World
Bank.
Cuddington J.and J. Hancock (1994). “Assessing the impact of AIDS on
the growth path of the Malawian economy”. Journal of Development
Economics, 43:363-368.
Daily Nation (2001). AIDS campaign setback: many Kenyans stick to
their sexual habits, says survey. 24 September 2001. Nairobi:
Nation Media.
Daily Nation (2003). New push for mass AIDS testing: compulsory HIV/
screening a good plan. Horizon, Daily Nation, 20 March 2003.
Nairobi: Nation Media.
Denny M. and M. Fuss (1977). The use of approximation analysis to
test for separability and the existence of consistent aggregates.
The America Economic Review, 67(3): 404-418.
Douglas P.H. (1976). The Cobb-Douglas production function once again:
its history, testing and some new empirical values. Journal of
Political Economy, 84(5): 903-91.
Family Health International/UNAIDS (2001). Effective prevention
strategies in low prevalence settings. Family Health International,
USA.
Family Health International/AIDSCAP (1996). AIDS in Kenya:
Socioeconomic impact and policy implications. Family Health
International, USA.

64
 References

Forsythe, S., D. Sokal, L. Lux, T.D.N. King (1993). An assessment of the

economic impact of AIDS in Kenya.” Family Health International
AIDSTECH/AIDSCAP.
Forsythe, S., M. Roberts (1995). “The financial impact of HIV/AIDS on
Africa’s commercial sector: A case study from Kenya.”
AIDScaptions, Vol. 2, No.1.
GoK (Government of Kenya) (1994). Kenya’s health policy framework.
Nairobi: Government Printer.
GoK (Government of Kenya) (1997). Sessional Paper No. 4 of 1997 on AIDS
in Kenya. Nairobi: Ministry of Health.
GoK (Government of Kenya) (2001a. Economic Survey, 2001. Nairobi:
Government Printer.
GoK (Government of Kenya) (2001c). National condom policy and strategy.
Nairobi: Ministry of Health and National AIDS Control Council.
GoK (Government of Kenya) (2003). Economic Survey, 2003. Nairobi:
Government Printer.
GoK (Government of Kenya)/UNICEF (2003). Multiple indicator cluster
survey (MICS) Report, 2000. Nairobi: Central Bureau of Statistics
and Ministry of Planning and National Development.
Green E.C. (2002). What happened in Uganda? What are the lessons
from Uganda for AIDS prevention? Presentation to USAID,
Washington DC, 5 February 2002.
Greener R., K. Jeffris and H. Siphambe (2000). The impact of HIV/AIDS
on poverty and inequality in Botswana. The South African Journal
of Economics, 68(5): 888-915.
IEA (Institute of Economic Affairs)/(SID) Society for International
Development (2001). Kenya at the crossroads: scenarios for our future.
Nairobi: IEA and SID.
KIE (Kenya Institute f Education ) (1997). “AIDS education for youth
action programme” Facilitator ’s handbook, Nairobi: AIDS
education project. Nairobi: Kenya Institute of Education.
Leighton, C. (1993). “Economic and social impacts of the HIV/AIDS
epidemic in African and Asian settings: Case studies of Kenya
and Thailand” Abt Associates.
Lonwenson R. and A. Whiteside (2001). HIV/AIDS: implications for
poverty reduction. Background paper prepared for the UNDP
for the General Assembly, Special Session on HIV/AIDS, 25-27
June 2001.

65
HIV/AIDS in Kenya: A review of research and policy issues

Lundberg M., M. Over and P. Mujinja (2000). Sources of financial

assistance for households suffering and adult death in Kagera,
Tanzania. The South African Journal of Economics 68(5):947-984.
Malaney P. (2000). Microeconomic approaches to evaluating the burden
of malaria. Cambridge, Mass.: Harvard University, Center for
International Development. Mimeo.
Marum E. and E. Madraa (1999). A decade of an effective national
response to AIDS: a review of the Ugandan experience (Draft).
Morris C. N., D. R. Burdge and E.J. Cheevers (2000). Economic impact
of HIV infection in a cohort of male sugar mill workers in South
Africa. The South African Journal of Economics, 68 (5): 933-946.
NACC (National Aids Control Council) (2002). Mainstreaming gender
into the Kenya National HIV/AIDS Strategic Plan 2002-2005. The
Gender and HIV/AIDS Technical Sub-committee of the National
AIDS Control Council.
NACC (National Aids Control Council) (2000). The Kenya National HIV/
AIDS Strategic Plan 2000-2005. Popular Version. Nairobi: NACC.
Nalo D. and M. Aoko (1993). “Economic impact of AIDS in Kenya.”
Nairobi: Ministry of Planning and Development, Long Range
Planning Division.
NASCOP (1994: AIDS in Kenya: background projections, impact,
interventions, policy. Nairobi: Ministry of Health.
NASCOP (1998): AIDS in Kenya: background projections, impact,
interventions, policy. Nairobi: Ministry of Health.
NASCOP (1999): AIDS in Kenya: background projections, impact,
interventions, policy. Nairobi: Ministry of Health.
Nicholls S., R. Mc Lean, K. Theodore, R. Henry, B. Camara (2000).
Modeling the macroeconomic impact of HIV/AIDS in the English
speaking Caribbean. The South African Journal of Economics 68(5):
916-932.
Rugalema G., S. Weigang and J. Mbwika (1998). HIV/AIDS and the
commercial agricultural sector of Kenya: impact, vulnerability,
susceptibility and coping strategies. Consultancy Report KE95/
100. Rome: FAO.
Steinberg M., A. Kinghorn, N. Soderlund, G. Schierhout, S. Conway,
Abt Associates (2000). HIV/AIDS facts figures and the future:
capacity of traditional coping mechanisms. SARH 2000.
Stover J. (2002). The effects of behaviour change on trends in HIV
incidence in Uganda and Kenya. Presentation to USAID,
Washington DC, 5 February 2002.

66
 References

UNAIDS (2002). Epidemiological fact sheet: 2002 update Kenya.

Geneva: UNAIDS.
UNAIDS (2000). “HIV and healthcare reform in Phayao: From crisis to
opportunity”. Geneva: UNAIDS.
UNAIDS (1995). Report from a consultation on the socio-economic
impact of HIV/AIDS on households. UNAIDS.
UNAIDS/WHO (2000). Epidemiological fact sheet on HIV/AIDS and
sexually transmitted infections – Kenya. Geneva: UNAIDS.
WHO (World Health Organization)/UNAIDS (2002). Epidemiological fact
sheets on HIV/AIDS and sexually transmitted infections: Kenya.
Geneva: WHO.
World Bank (1997). Confronting AIDS: Public priorities in a global
epidemic. New York: Oxford University Press.

67
HIV/AIDS in Kenya: A review of research and policy issues

Appendix
Table A1: Estimated number of people living with HIV/AIDS
worldwide, 2002
Region Adults and Children % women
children (<15)

Sub-Saharan Africa 29,400,000 2,800,000 58

East Asia and Pacific 1,200,000 4,000 24
Australia & New Zealand 15,000 <200 7
South and Southeast Asia 6,000,000 240,000 36
Eastern Europe and central Asia 1,200,000 15,000 27
Western Europe 570,000 5,000 25
North Africa and Middle East 550,000 40,000 55
North America 980,000 10,000 20
Caribbean 440,000 20,000 50
Latin America 1,500,000 45,000 30

Total 42,000,000 3,200,000 50

WHO/UNAIDS (2002)

Table A2: Estimated number of new infections and deaths from HIV/
AIDS, 2002
Region New HIV infections, 2002 Deaths from HIV/AIDS, 2002

Adults & Adults Children Adults & Adults Children

children (<15) children (<15)
Sub-Saharan
Africa 3,500,000 2,700,000 720,000 2,400,000 1,800,000 550,000
East Asia and
Pacific 270,000 270,000 3,000 45,000 43,000 2,000
Australia and
New Zealand 500 500 <100 <100 <100 <100
South and South
East Asia 700,000 640,000 60,000 440,000 400,000 43,000
Eastern Europe
and central Asia 250,000 250,000 1,000 25,000 25,000 <100
Western Europe 30,000 30,000 <500 8,000 8,000 <100
North Africa and
Middle East 83,000 70,000 13,000 37,000 30,000 6,800
North America 45,000 45,000 <500 15,000 15,000 <100
Caribbean 60,000 50,000 7,000 42,000 35,000 7,000
Latin America 150,000 140,000 10,000 60,000 55,000 5,000

Total 5,000,000 4,200,000 800,000 3,100,000 2,500,000 610,000

WHO/UNAIDS (2002)

68
 Appendix

Table A3: Regional distribution of HIV, 2002

Province Urban Rural Total

Number % Number % Number %

Central 32,113 7 204,383 14 236,496 12

Coast 61,509 13 71,771 5 133,280 7
Eastern 47,823 10 330,316 23 378,139 19
Nairobi 174,747 36 - - 174,747 9
Northeastern 2,620 1 12,188 1 14,808 1
Nyanza 63,681 13 382,712 26 446,393 23
Rift Valley 80,989 17 285,431 19 366,419 19
Western 16,194 3 178,884 12 195,080 10
National 481,364 100 1,465,685 100 1,945,363 100

Source: Government of Kenya (2003)

Table A4: HIV prevalence rates among STI patients by sentinel year
and marital status

Year Monogamous Polygamous Separated Single Widowed

1990 13.9 0.0 34.5 15.3 16.7
1991 22.0 0.0 40.9 20.9 16.7
1992 28.9 0.0 41.7 29.8 50.0
1993 21.8 25.0 35.7 25.4 37.5
1994 25.7 13.9 28.6 28.4 42.9
1995 25.8 0.0 0.0 25.0 0.0
1996 21.5 30.9 20.7 20.2 50.0
1997 39.2 54.5 65.7 38.1 73.9
1998 30.9 48.8 68.8 23.0 50.0
1999 28.6 66.7 25.0 23.4 33.3
2000 31.2 54.0 40.0 30.1 68.2
2001 25.5 37.1 40.2 20.3 66.7

Source: Government of Kenya (2003)

69
HIV/AIDS in Kenya: A review of research and policy issues

Table A5: Percentage of men and women aware of methods of preventing

HIV transmission, 2000
Category Women Men
Have one Use of Knows Have one Use of Knows

Abstaining

Abstaining
faithful a at least faithful a at least
uninfected condom one uninfected condom one
partner every way partner every way
time time
Province
Nairobi 78.6 61.6 78.4 85.7 86.0 62.7 86.7 91.7
Central 81.9 56.2 77.9 83.4 94.5 71.4 88.7 95.7
Coast 78.0 56.1 75.6 81.6 89.5 71.1 88.7 93.2
Eastern 81.8 46.9 77.6 84.6 93.2 65.7 87.0 95.4
Northeastern 53.2 29.0 51.6 53.7 75.4 38.7 71.5 76.9
Nyanza 77.4 45.7 70.9 84.5 89.1 62.2 82.2 95.7
Rift Valley 79.5 51.7 72.6 82.6 90.6 67.8 84.6 94.7
Western 76.0 49.9 71.7 81.4 85.6 69.6 81.3 91.1
Rural 79.0 49.6 73.6 82.9 90.7 66.8 84.5 94.3
Urban 78.7 59.0 77.6 84.4 86.9 65.4 87.4 92.3
Education
None 68.0 39.3 63.9 71.9 83.6 59.9 78.5 88.3
Primary 78.5 50.8 73.3 83.2 88.7 66.2 83.5 92.7
Secondary+ 85.9 62.2 83.6 89.9 93.0 68.8 90.0 97.4
Total 78.9 52.2 74.7 83.3 89.7 66.4 85.2 93.9
Source: Government of Kenya (2003); Government of Kenya/UNICEF (2003)

Table A6: Percentage of women and men who know of prevention

method, a place to take a test and who have been tested
Category Women Men
Have Use of Know a Have Have Use of Know a Have
Abstaining

Abstaining

one a place to been one a place to been

faithful condom be tested faithful condom be tested
uninfected every tested uninfected every tested
Age partner time partner time

15-19 70.7 47.3 66.6 50.3 4.8 80.0 62.6 75.1 56.8 2.0
20-24 80.1 57.8 76.7 65.9 10.9 87.6 68.5 84.2 75.7 4.7
25-29 83.3 57.7 78.1 73.5 14.2 91.7 73.8 86.5 80.8 10.2
30-34 80.6 52.6 77.0 69.6 15.7 93.4 71.4 88.7 84.0 12.9
35-39 80.5 48.4 76.0 65.0 9.7 94.8 63.2 89.0 85.7 9.6
40-44 82.0 52.5 77.5 59.8 10.7 94.8 69.1 91.9 84.5 10.0
45-49 79.2 38.9 73.5 62.4 5.5 90.1 62.1 88.6 79.4 11.6
Region
Rural 79.0 49.6 73.6 59.8 5.7 90.7 66.8 84.5 73.8 5.7
Urban 78.7 59.0 77.6 73.9 23.1 86.9 65.4 87.4 83.2 15.4
Education
None 68.0 39.3 63.9 47.4 6.3 83.6 59.9 78.5 64.5 8.7
Primary 78.5 50.8 73.3 60.8 8.1 88.7 66.2 83.5 72.5 5.2
Secondary+ 85.9 62.2 83.6 78.5 17.6 93.0 68.8 90.0 85.6 12.9
Total 78.9 52.2 74.7 63.6 10.4 89.7 66.4 85.2 76.3 8.2
Source: Government of Kenya (2003); Government of Kenya/UNICEF (2003)

70
 Appendix

Table A7: Percentage of women and men reporting discriminatory

behaviour against persons with HIV/AIDS, 2000

Category Women Men

Believe a Would not Agree with Believe a Would not Agree
teacher buy food at least one teacher buy food with at
with HIV from a of these with HIV from a least one
should not person should not person of these
teach with HIV/ teach with HIV/
AIDS AIDS
Province
Nairobi 49.5 35.1 56.5 62.5 40.6 66.3
Central 38.4 32.4 45.8 46.3 41.6 53.3
Coast 42.2 38.0 50.3 55.7 44.2 61.5
Eastern 35.2 29.7 41.8 44.8 37.8 52.2
Northeastern 28.3 22.7 32.5 58.2 56.7 63.3
Nyanza 37.2 38.5 48.6 50.2 45.4 60.7
Rift Valley 36.6 26.0 42.8 46.6 37.4 54.2
Western 37.9 32.4 47.2 54.8 46.3 63.2
Rural 35.7 30.9 43.9 46.1 39.8 54.4
Urban 49.0 36.7 55.9 65.1 46.3 69.7
Education
None 26.9 24.3 33.7 43.1 35.8 48.6
Primary 36.3 29.9 44.7 44.3 36.0 52.7
Secondary+ 52.7 42.5 59.9 64.2 51.9 70.2
Total 39.3 32.5 47.2 51.1 41.5 58.4

Source: Government of Kenya (2003); Government of Kenya/UNICEF (2003)

71
HIV/AIDS in Kenya: A review of research and policy issues

KIPPRA PUBLICATIONS
Conference Proceedings
Report of the proceedings of the AERC-KIPPRA World Trade Organization (WTO)
Workshop, 2000
Report of the proceedings of the International Conference on Finance and
Development: Evidence and Policy Issues, 2001

Discussion Papers
Njuguna S. Ndung’u (2000). The exchange rate and the interest rate differential in
Kenya: a monetary and fiscal policy dilemma. KIPPRA DP No. 1
Karingi, S. N. and Njuguna S. Ndung’u (2000). Macro models of the Kenyan
economy: a review. KIPPRA DP No. 2
Ronge, E. E. and H.O. Nyangito (2000). A review of Kenya’s current
industrialization policy. KIPPRA DP No. 3
Nyangito, H.O. (2001). Delivery of services to smallholder coffee farmers and
impacts on production under liberalization in Kenya. KIPPRA DP No. 4
Njuguna S. Ndungu and R. W. Ngugi (2000). Banking sector interest rate spread
in Kenya. KIPPRA DP No. 5
Karingi, S.N., M.S. Kimenyi and Njuguna S. Ndung’u (2001). Beer taxation in
Kenya: an assessment. KIPPRA DP No. 6
Ikiara, M.M. (2001).Vision and long term development strategy for Kenya’s
tourism industry. KIPPRA DP No. 7
Geda, A. and Njuguna S. Ndung’u (2001). Specifying and estimating partial
equilibrium models for use in macro models: a road map for the KIPPRA-
Treasury Macro Model. KIPPRA DP No. 8
Geda, A., Niek de Jong, G. Mwabu and M.S. Kimenyi (2001). Determinants of
poverty in Kenya: household-level analysis. KIPPRA DP No. 9
Were, M., A. Geda, S.N. Karingi and Njuguna S. Ndungu (2001). Kenya’s
exchange rate movement in a liberalized environment: an empirical analysis.
KIPPRA DP No. 10
Huizinga, F., A. Geda, Njuguna S. Ndung’u and S.N. Karingi
(2001).Theoretical base for the Kenya macro model: the KIPPRA-Treasury
macro model. KIPPRA DP No. 11
Mwabu, G., M. S. Kimenyi, P. Kimalu, N. Nafula and D. K. Manda (2002).
Predicting household poverty: a methodological note with a Kenyan example.
KIPPRA DP No. 12
Manda, D.K., G. Mwabu, M. S. Kimenyi (2002). Human capital externalities and
returns to education in Kenya. KIPPRA DP No. 13
Bedi, A., P.K. Kimalu, D.K. Manda, N.N. Nafula (2002).The decline in primary
school enrolment in Kenya. KIPPRA DP No. 14

72
Odhiambo, W. and H. Nyangito (2002). Land laws and land use in Kenya:
implications for agricultural development. DP No. 15
Were, M. and S. Karingi (2002). Better understanding of the Kenyan economy:
simulations from the KIPPRA-Treasury Macro Model. KIPPRA DP No. 16
Nyangito, H., M. Ikiara and E. Ronge (2002). Performance of Kenya’s wheat
industry and prospects for regional trade in wheat products. DP No. 17
Nyangito, H. and L. Ndirangu (2002). Impact of institutional and regulatory
framework on the food crops subsector in Kenya: 1990-1999. KIPPRA DP No.
18
Ikiara, M. (2002). Impact of tourism on environment in Kenya: status and policy.
KIPPRA DP No. 19
Ronge, E., L. Ndirangu and H. Nyangito (2002). Review of government policies
for the promotion of micro and smallscale enterprises in Kenya. KIPPRA DP.
No. 20
Kiringai, J., Njuguna S. Ndung’u, and S.N. Karingi (2002).Tobacco excise tax
in Kenya: an appraisal. KIPPRA DP No. 21
Were, M., Njuguna S. Ndung’u, A. Geda and S.N. Karingi (2002). Analysis of
Kenya’s export performance: an empirical evaluation. KIPPRA DP No. 22
Mwangi S. Kimenyi (2002). Ethnicity, institutions og governance and conflict
avoidance. KIPPRA DP No. 23
Ikiara, M.M., L. Ndirangu (2003).Prospects of Kenya’s clothing exports under
AGOA after 2004. KIPPRA DP No. 24
Nyangito, H. (2003). Agricultural trade reforms in Kenya under the WTO
framework. KIPPRA DP No. 25
Odhiambo, W. and H. Nyangito (2003). Measuring agricultural productivity in
Kenya: a review of approaches. KIPPRA DP No. 26
Ngugi, R.W. (2003). Development of the Nairobi Stock Exchange: a historical
perspective. KIPPRA DP No. 27
Njuguna, A. E., S.N. Karingi and M.S. Kimenyi (2003). Alternative
methodologies for measuring Kenya’s potential output and output gap. KIPPRA
DP No. 28
Ngugi, R. W. (2003). What defines liquidity of the stock market? The case of the
Nairobi Stock Exchange. KIPPRA DP. No. 29
Nafula, N.N. (2003). Bank portfolios and bank earnings in Kenya: an econometric
analysis. KIPPRA DP No. 30.
Manda, D.K. (2004). Globalisation and the labour market in Kenya. KIPPRA DP.
No. 31.
Bedi, A., P. Kimalu, M.S. Kimenyi, D.K. Manda, G. Mwabu and N. Nafula
(2004). User charges and utilisation of health services in Kenya. KIPPRA DP.
No. 32.

73
HIV/AIDS in Kenya: A review of research and policy issues

Oiro, M. W., G. Mwabu and D.K. Manda (2004). Poverty and employment in
Kenya. KIPPRA DP No. 33
Odhiambo, W., H. O. Nyangito and J. Nzuma (2004). Sources and determinants
of agricultural productivity in Kenya. KIPPRA DP No. 34
Muthaka, David I, Diana N. Kimani, Stephen Mwaura, Damiano K. Manda
(2004). A review of the regulatory framework for private healthcare services in
Kenya. KIPPRA DP No. 35
Kamau, A., S. Karingi, Njuguna S. Ndung’u, S. Mwaura. Capital requirements
and bank behaviour in Kenya: Empiical evidence. KIPPRA DP No. 36
Nafula, N. N., P. K. Kimalu, J. Kiringai, R. Owino, D. K. Manda, S. Karingi.
Budget mechanisms and public expenditure tracking in Kenya. KIPPRA DP
No. 37.

Occassional Papers
Gitu, K. W. (2001). Strengthening the link between policy research and
implementation. KIPPRA OP No. 1
Kimenyi, M.S. (2001). Effective private sector representation in policy formulation
and implementation. KIPPRA OP No. 2
Kimenyi, M.S. (2002). Agriculture, economic growth and poverty reduction.
KIPPRA OP No. 3
Nyangito, H. (2002). Post-Doha African challenges in the sanitary and
phytosanitary and trade related intellectual property rights agreement. KIPPRA
OP No. 4
Mwabu, G. (2004). Principles of research. KIPPRA OP No. 5

Policy Papers
Nyangito, H.O. (2001). Policy and legal framework for the tea subsector and the
impact of liberalization in Kenya. KIPPRA PP No. 1
Nyangito, H.O. (2001). Policy and legal framework for the coffee subsector and the
impact of liberalization in Kenya. KIPPRA PP No. 2
Ikiara, M.M. and H. Nyangito (2001). Effects of visa waiver and increase in airport
tax on Kenya’s tourism industry. KIPPRA PP No. 3

Special Reports
Legal and other constraints on access to financial services in Kenya: survey results.
KIPPRA Private Sector Development Division. SR No. 1, 2001
Thinking about regulating? The better regulation guide. KIPPRA Private Sector
Development Division. SR No. 2, 2002
Policy timeline and time series data for Kenya: an analytical data compendium. KIPPRA
Macroeconomics Division, SR No. 3, 2002

74
Tax analysis and revenue forecasting in Kenya. KIPPRA Macroeconomics Division,
SR No. 4, 2003
Data compendium for Kenya’s agricultural sector. KIPPRA Productive Sector
Division, SR No. 5, 2003

Working Papers
Wasike, W.S.K. (2001). Road infrastructure policies in Kenya: historical trends and
current challenges. KIPPRA WP No. 1
Ikiara, M.M. (2001). Policy framework of Kenya’s tourism sector since independence
and emerging policy concerns. KIPPRA WP No. 2
Manda, D.K., M.S. Kimenyi and G. Mwabu. A review of poverty and antipoverty
initiatives in Kenya. KIPPRA WP No. 3
Kimalu, P.K., N. Nafula, D.K. Manda, G. Mwabu and M.S. Kimenyi (2001).
Education indicators in Kenya. KIPPRA WP No. 4
Geda, A., S.N. Karingi, Njuguna S. Ndung’u, M. van Schaaijk, M. Were, W.
Wassala and J. Obere (2001). Estimation procedure and estimated results of the
KIPPRA-Treasury macro model. KIPPRA WP No. 5
Kimalu, P., N. Nafula, D.K. Manda, G. Mwabu and M.S. Kimenyi (2002). A
situational analysis of poverty in Kenya. KIPPRA WP No. 6
Kiringai, J. and G. West (2002). Budget reforms and the Medium-Term Expenditure
Framework in Kenya. KIPPRA WP No. 7
Ikiara, M. and L. Ndirangu (2003). Developing a revival strategy for Kenya’s cotton-
textile industry: a value chain approach. KIPPRA WP No. 8
Ng’eno, N.K., H.O. Nyangito, M.M. Ikiara, E.E. Ronge, J. Nyamunga (2003).
Regional integration study of East Africa: the case of Kenya. KIPPRA WP No. 9
Manda, D. K., P.K. Kimalu, N. Nafula, Diana K. Kimani, R. K. Nyaga, J.M.
Mutua, G. Mwabu, M.S. Kimenyi (2003). Cost and benefits of eliminating child
labour in Kenya. KIPPRA Working Paper No. 10
Kimalu, P.K., N.N. Nafula, D.K. Manda, A. Bedi, G. Mwabu, M.S. Kimenyi
(2004). A review of the health sector in Kenya. KIPPRA Working Paper No. 11

75
 HIV/AIDS in Kenya: A review of research and policy issues

76

View publication stats